(PDF) Quantitative, Qualitative and Mixed Method Research Methodology

Quantitative,

Qualitative

and Mixed Method

Research Methodology

Jonathan Sarwono

ISBN: 9781076287885

ISBN-13: 9781076287885.

Published by: Amazon.com, Inc. 410 Terry Avenue North Seattle, Washington

98109 US

DEDICATION

This book is dedicated to Chloe Andrea and Regina Tiatira

Table of Contents

Foreword

1

PRELIMINARY

1

2

RESEARCH DEFINITION AND TYPES

8

3

RESEARCH PROPOSAL

11

4

THE STAGES OF RESEARCH PROCES

14

5

CHOOSING PROBLEMS AND MAKING HYPOTHESES

17

6

STUDY OF LITERATURE

25

7

IDENTIFYING AND GIVING VARIABLE NAMES

28

8

DEVELOPING OPERATIONAL VARIABLE DEFINITIONS

38

9

TECHNIQUES FOR MANIPULATING AND CONTROLLING

VARIABLES

40

10

MAKING RESEARCH DESIGN

44

11

MEASUREMENT SCALE

53

12

DATA COLLECTION INSTRUMENT DEVELOPMENT

58

13

SAMPLE DESIGN

63

14

DATA COLLECTION

70

15

PROCESSING AND ANALYSIS OF DATA

76

16

REPORT WRITING

80

17

QUALITATIVE RESEARCH DEFINITION

84

18

THE FUNDAMENTAL THEORIES

86

19

QUALITATIVE RESEARCH DESIGN

87

20

INFORMANT RECRUITMENT TECHNIQUES

90

21

DATA TYPES

92

Quantitative, Qualitative and Mixed Method Research Methodology

v

22

DATA COLLECTION INSTRUMENT

93

23

DATA COLLECTION METHOD

96

24

ANALYSIS TECHNIQUES

104

25

VALIDITY OF QUALITATIVE RESEARCH

110

26

PRESENT QUALITATIVE RESEARCH RESULTS

112

27

WRITE A REPORT

116

28

COMBINING QUANTITATIVE AND QUALITATIVE

APPROACHES

118

ACKNOWLEDGMENTS

This book is about how to use quantitative and qualitative research and how to

combine the methods among those approaches. It starts from philosophical

common grounds, theory, and application. There is a tendency of current

researchers using the combined methods in which people regard it more

satisfactorily. Above from pros and cons the fact is that it is not easy to use the

combined methods correctly. This book tries to suggest some models, hopefully it

can help readers to use each approach and method correctly.

Suggestions and questions can be sent to the following email:

jsarwono007@gmail.com or jonathan@iwu.ac.id

Bandung, April 2022

Author

1

CHAPTER I

PRELIMINARY

.

1.1 Human Truth Searching

In everyday life since ancient times humans have always tried to find the

essence of truth about things that are essential, such as God's problems, death,

life after death, love and others. Humans try to understand and conquer a

universe full of mysteries. Until the era that was colored with the

sophistication of technology today, the feeling to understand and understand

the secrets of the universe including secrets about himself.

In the middle ages, humans have not shown interest in systematic studies of

the physical world, these conditions are much influenced by Greek

philosophical opinions which prioritize "the general" rather than "the special".

General knowledge refers to the essence and essence of concrete things, while

special things distinguish between one thing and another.

In Greek mythology the term god Zeus is known to be associated with

weather, rain and lightning, the god Poseidon who ruled the oceans and

earthquakes. When natural disasters such as earthquakes, floods and others

occur; humans always connect with supernatural things. In the development of

his thinking finally humans after experiencing various processes succeed in

using the inner reasoning power (ratio) solve the problem. As happened in the

Middle Ages with scientific discoveries by Copernicus and Edison. As the

opinion of a philosopher Rene Descartes who said " Cogito Ergo Sum " (I was

there because of thinking) then humans began to use his extraordinary mind.

Even so, it is necessary to distinguish between the use of common sense and

science. The basic difference between the two is the word "systematic" and

"controlled". There are five main things that distinguish between science and

common sense. First, science is developed through theoretical structures, and

tested for internal consistency. In developing its structure, it is done by

testing or testing empirically. While the use of common sense is usually not.

Jonathan Sarwono

2

Secondly, in science, theories and hypotheses are always tested empirical.

This is with people who are not scientists in a "selective" way. Third, there is

an understanding of control which in scientific research can have various

meanings. Fourth, science emphasizes the existence of a connection between

phenomena consciously and systematically. The pattern of connecting is not

done carelessly. The fifth, the difference lies in how to give different

explanations in observing a phenomenon. In explaining the relationship

between phenomena, scientists carry out caution and avoid metaphysical

interpretations. The resulting propositions are always open for scientific

observation and testing.

1.1 The Occurrence of the Natural Secularization Process

In the beginning, humans considered nature to be sacred, so that between

subjects and objects there is no limit. In its development as mentioned above,

there has been a shift in the concept of law (nature). Law is defined as fixed

links and must be between symptoms. Regular links in nature have always

been interpreted into normative laws. Here this understanding is associated

with God or para god as the creator of the law that must be obeyed. Towards

the 16th century humans began to abandon the notion of normative law.

Instead the understanding of the law appears in accordance with natural law.

This understanding implies that there is an order in nature and the order can

be concluded through empirical research. Scientists at that time argued that

God as the creator of natural law gradually gained abstract and impersonal

nature. Nature has lost its sacredness, instead a picture of the world that

corresponds to the natural sciences for modern humans with human

scientific abilities begins to open up the secrets of nature.

1.2 Various Ways to Find the Truth

In human history, efforts to search for truth have been done in various ways

such as:

Coincidentally: There is a story whose truth is hard to trace regarding the

discovery of a malaria drug case that happened by accident. When an Indian

is sick and drinks water in the pond and finally gets a cure. And that happens

repeatedly in some people. Finally, it was discovered that around the ponds

grew a kind of tree whose skin was used as a malaria medicine which then fell

on the pond. The discovery of the tree that would later be known as the

quinine tree was accidental.

Trial And Error: Another way to get the truth is to use the "trial and error"

method which means trial and error. This method is chancy. One example is

Quantitative, Qualitative and Mixed Method Research Methodology

3

the experimental model of the "problem box" by Thorndike. The experiment is as

follows: a hungry cat is put into a "problem box" - a room that can only be

opened when the cat manages to pull the end of the rope by opening the

door. Because of hunger and seeing food outside the cat trying to get out of

the box in various ways. Finally, accidentally the cat managed to touch the

knot of the rope that made the door open and he managed to get out. The

experiment was based on uncertain things, namely the ability of the cat to

open the door to the problem box.

Use of Authority: Truth can be obtained through the authority of someone

who holds power, such as a king or a government official whose decisions

and policies are considered right by his subordinates. In Javanese philosophy

known as 'Sabita pendita ratu', it means that the words of the king or priest

are always true and cannot be denied again.

Problem Solving by Speculation: Solving the problem with the "trial and

error" method which emphasizes the element of chancy and uncertainty and

accuracy.

Critical Thinking / Based on Experience: Another method is thinking

critically and based on experience. An example of this method is thinking

deductively and inductively. Deductively means to think from the general to

the special; being inductive from being specific to the general one. Deductive

methods have been used for hundreds of years since Aristotle's time.

Use of Scientific Investigation: According to Francis Bacon a new truth

can be obtained by using scientific inquiry, critical and inductive thinking.

Problem Solving Method: The problem-solving method developed by

Karl. R. Popper in 1937 was a variation of the "trial and error" method . this

method shows the scheme as follows: P1-TS-EE-P2, P1 is the initial

problem, TS tentative solution - the theory that is attempted to submit, EE is

"error elimination" - evaluation with the aim of finding and removing errors,

and P2 is a new situation caused by a critical evaluation of the solution

tentative to the initial problem so that new problems arise.

Knowledge Basics: In this section, the basics of knowledge that will

spearhead scientific thinking will be discussed. The basics of knowledge are

as follows:

Jonathan Sarwono

4

Reasoning: What is meant by reasoning is the activity of thinking according to

a certain pattern, according to certain logic in order to produce knowledge.

Logical thinking has multiple connotations, is analytical. This stream that uses

reasoning as a source of truth is called the flow of rationalism and which

considers facts to be captured through experience as truth called the

empiricism flow.

Logic (How to withdraw conclusions): The second characteristic is logic or

how to draw conclusions. What is meant by logic as defined by William SS is

"study to think validly. In logic there are two kinds, namely inductive and

deductive logic.

Source of Knowledge: The source of knowledge in this world stems from

human attitudes that doubt every symptom in the universe. Humans do not

want to accept things that exist including the fate of himself. Rene Descarte

once said "de omnibus dubitandum" which means that everything must be

doubted. The issue of the criteria for establishing truth is hard to believe. From

various streams, various criteria of truth emerged.

Truth Criteria: One criterion of truth is the existence of consistency with

previous statements that are considered true. Some truth criteria include:

Coherence Theory: What is meant by the theory of coherence is that a

statement is considered true if the statement is coherent and consistent with

previous statements that are considered true. An example is mathematics,

which is a form of compilation, proving it based on coherent theory.

Correspondence Theory: The correspondence theory was pioneered by

Bertrand Russel. In this theory a statement is considered true if the material of

knowledge is conceived corresponds to the object intended by the statement.

An example is if someone says that the British capital is London, then that

statement is true. Whereas if he said that the capital of England was Jakarta,

then the statement was wrong; because in reality the British capital is London

not Jakarta.

Pragmatic Theory: The main character in this theory is Charles S Pierce.

Pragmatic theory says that the truth of a statement is measured by the criteria

whether the statement is functional in practical life. The truth criteria are based

on the usefulness of the theory. Besides that this school believes that a theory

will not last , in a certain period of time it can be changed by making a revision.

Quantitative, Qualitative and Mixed Method Research Methodology

5

Ontology: Ontology is the essence of what is studied or the knowledge itself. A

philosopher named Democritus explained the principles of materialism to say as

follows: Just based on habit, sweetness is sweet, the heat is hot, cold is cold, the

color is color. That is, sensing objects are often considered real, but not so. Only

the atom and void are real. So the term " sweet , hot and cold" is only the

terminology that we give to the symptoms captured by the senses.

Science is knowledge that tries to interpret this universe as it is, therefore humans

in exploring knowledge cannot be separated from the symptoms that are in it. And

the nature of science that serves to help humans in solving problems does not

need to have absolutes such as religion that provide guidance on the most essential

things from this life. Even so, to a certain extent, science needs to have validity in

making generalizations. For example, how we define humans, then various

meanings will emerge.

Epistemology: Epistemology is meant to get the right knowledge. Some things

that need to be considered in getting knowledge are: Limitation of the study of

science: ontologically limiting science to the study of objects that are within the

scope of humans. Cannot review transcendental areas. How to compile knowledge:

to get knowledge into science a way is needed to compile it by using the scientific

method. 3. Required foundation that is in accordance with the ontological and

axiological science itself. Explanation directed at the description of the relationship

of various factors that are bound in a constellation of causes of the emergence of a

symptom and the process of occurrence. 4. The scientific method must be

systematic and explicit. 5. The scientific method cannot be applied to knowledge

that does not belong to the group of knowledge. 6. Science tries to find an

explanation of nature and makes conclusions that are general and impersonal.

7.Characteristics that stand out the theoretical framework: Exact sciences:

deductive, ratio, quantitative. Social Sciences: inductive, empirical, qualitative

Some Basics Concept: Concepts are terms and definitions used to describe

symptoms in the abstract, for example events, circumstances, groups. It is expected

that the researcher is able to formulate his thoughts into concepts clearly in

relation to the simplification of several problems related to each other. In the

world of research there are two notions of concepts, namely: First, concepts that

are clearly related to reality are represented, for example: tables, cars etc. Second,

concepts that abstract their relationship with reality are represented, for example:

intelligence, kinship, etc.

Construct: The construct is a concept created and used with intent and awareness

for specific scientific purposes.

Proposition: Proposition is a logical relationship between two concepts. In social

research there are two types of propositions known; the first axiom or postulate,

Jonathan Sarwono

6

the second theorem. Axioms are propositions whose truth is no longer in

research, while theorems are propositions which are deduced from axioms.

Theory: One definition of theory is a series of assumptions, concepts,

constructs, definitions and propositions to explain a phenomenon

systematically by way of formulating relationships between concepts. Shone

(2022) said that “A theory is a generalized synthetic explanatory statement, in

other words, an abstract conceptual explanation of the world. Conceptual

explanations are important in order to generalize our findings across a wide

range of situations, but unless they lead to predictions, we cannot test our

theories, so we need models” Another definition says that theory is scientific

knowledge that includes an explanation of a particular invoice from one

scientific discipline.

Theory has several characteristics as follows: a) must be consistent with

previous theories allows no contraction in overall scientific theory; b) must

be compatible with empirical facts, because the theory which, however

consistent, if not supported by empirical testing is not scientifically

acceptable.

There are four ways the theory is built according to Melvin Marx: 1) Model

Based Theory, 2) Deductive Theory, 3) Inductive Theory, and 4) Functional

Theory. Based on the first theory the theory developed with the existence of

conceptual networks which were then tested empirically. The substance

validity lies in the initial stages of testing the model, namely whether the

model works according to the needs of the researcher.

The second theory says a theory is developed through a process of

deduction. Deduction is a form of inference which decreases a conclusion

obtained through the use of logical thought accompanied by premise as

evidence. Deductive theory is a theory that emphasizes conceptual structure

and substance validity. This theory also focuses on building concepts before

empirical testing.

The third theory emphasizes the empirical approach to get generalizations.

Drawing conclusions is based on repeated reality observations and develops

statements that function to explain and explain the existence of these

statements.

The fourth theory says a theory is developed through continuous interaction

between the conceptualization process and empirical testing that follows it.

The main difference with deductive theory lies in the process of

conceptualization at the beginning of the development of the theory. In

Quantitative, Qualitative and Mixed Method Research Methodology

7

deductive theory the design of conceptual relationships is formulated and

testing is carried out at the final stage of the development of the theory.

Scientific Logic: The combination of deductive and inductive logic where

rationalism and empiricism are together in a system with a corrective

mechanism.

Hypothesis: The hypothesis is a temporary answer to the problem being

studied. The hypothesis is a suggestion of scientific research because a

hypothesis is a work instrument of a theory and is specific in nature that is

ready to be tested empirically. In formulating the hypothesis the statement

must be a reflection of the relationship between two or more variables.

Relational or descriptive hypotheses are called work hypotheses (H1), while

statistical tests require a work hypothesis and a reverse formulation of the

working hypothesis. Such a hypothesis is called the null hypothesis (H0).

Shone (2022) states hypothesis as a statement that can be empirically tested,

i.e. translation of theory into a testable statement

Variable: Variables are constructs or traits that are being studied. According

to Shone (2022) “A variable is something that varies over time or over

subjects (in other words, varies within the study), also used to mean the

operational definition of a concept (how do we measure something)”.

There are five types of variables that are known in the research, namely: the

independent variable, the dependent variable, moderate, intervening variables

and control variables. If viewed from the side of the measurement scale,

there are four types of variables: nominal, ordinal, interval and ratio.

Operational definition: What is meant by operational definition is the

specification of the activities of the researcher in measuring or manipulating a

variable. Operational definitions limit or mean a variable by detailing what

the researcher must do to measure the variable.

Means of Scientific Thinking: Language, Mathematics and Statistics

Axiology: Axiology is a matter of the value of the usefulness of science.

Science is not value free. This means that at certain stages sometimes science

must be adapted to the cultural and moral values of a society; so that the value

of the use of knowledge can be felt by the community in their efforts to

improve shared prosperity, not the other way around, instead it causes

disasters

Jonathan Sarwono

8

CHAPTER 2

RESEARCH DEFINITION AND TYPES

2.1 Definition

There are some definitions of the research. Shone (2022) define research as” A

systematic and unbiased way of solving a problem (by answering questions or

supporting hypotheses) through generating verifiable data.”. Another definition

states that “Research attempts to provide answers to questions. Such answers may

be very abstract and generally the case of basic research or they may be highly

concrete and specific as is often the case in applied research. (Tuckman 1978)

Based on the above definition simply can be said that research is systematic ways

to answer the problem being studied. Systematic words are keywords that are

related to the scientific method which means there is a procedure that is

characterized by order and completeness. In more detail Davis (1985) gives the

characteristics of a scientific method as follows:

First

The method must be critical, analytical, meaning that the method show

the right and correct process for identify problems and determine

methods for problem solution

Second

The method must be logical, meaning that there is a method used for

give away argumentation scientific. Conclusions made

rationally are based on available evidence.

Third

The method is objective, meaning that objectivity results investigations

that can be emulated by other scientists in the same study with the

same conditions

Fourth

Methods must be conceptual and theoretical; therefore that, to direct

the research process that is carried out researchers need to develop

concepts and structures theory so that the results can be accounted

for scientific view point.

Fifth

The method is empirical, meaning the method used based on facts in

the field

2.2 Types of Research

The types of research are distinguished based on the type of data needed in general

divided into two, namely primary research and secondary research.

2.2.1 Primary Research

Primary research requires data or information from the first source, we usually call

the respondent. Data or information is obtained through written questions using a

questionnaire or oral using the interview method. Included in this category are: a.

Case study: Case studies use individuals or groups as their study material. Usually

case studies are longitudinal. b. Survey: Surveys are quantitative studies that are

used to examine the symptoms of a group or individual behavior. In general,

Quantitative, Qualitative and Mixed Method Research Methodology

9

surveys use questionnaires as data collection tools. The survey adheres to the rules

of the quantitative approach, namely the larger the sample, the more it reflects the

resulting population. c. Experimental Research: Experimental research uses

individuals or groups as study material. In general, this research uses two or more

groups to serve as the object of their study. The first group is the group under

study while the second group is the comparison group (control group).

Experimental research uses a standard, structured and specific design.

2.2.2 Secondary Research

Secondary research uses material that is not from the first source as a means to

obtain data or information to answer the problem under study. This research is

also known as research that uses literature studies and which are usually used by

researchers who adhere to a qualitative approach. Suharsini Arikunto (1992)

divides the types of research based on a) purpose, b) approach, c) field of science,

d) place or setting, e) presence of variables.

2.2.3 Research is seen from its purpose

If research is seen from its objectives, then there are three sub-types of research,

namely exploratory research, verification research and development research.

Exploratory type research is used to conduct answers to why certain events arise,

for example the emergence of natural disasters in certain areas continuously.

Verification research is used to re-examine the results of previous research with the

aim of verifying the truth of the results of previous studies. Development research

aims to develop innovative models or things. This type of research is usually

carried out in a company in order to develop new products or services.

2.2.4 Research is seen from the approach

Viewed from the approach, this research is divided into two, namely longitudinal

and cross-sectional approaches. Approach first doing research based on a certain

period of time, usually a long time, for example a researcher conducts research on

the development of children's speaking abilities from the age of 10 months to 24

months. On the contrary, the second approach of the researchers conducted a

study of children's speaking abilities ranging from 10 months to 24 months

simultaneously at the same time.

2.2.5 Research is seen from the field of science

In this perspective, the types of research are divided based on their respective

disciplines, such as education research, technical research, economic research, etc.

2.2.6 Research seen from the location / background

When viewed from a place or setting where a researcher conducts research, then

this type of research is divided into three, namely: a) laboratory research, b) field

research, c) library research. Laboratory research is usually carried out in the exact

sciences, such as medical, electrical, civil, etc. Field research is usually carried out

by social and economic scientists where the location of the research is in a

particular community or group of people or certain objects as the setting where the

Jonathan Sarwono

10

researcher conducts research. Library research is carried out in the library by

reviewing the literature, previous research, journals and other sources in the library.

With the increasingly sophisticated information technology, this type of research at

this time does not have to be done in a physical library, but it can also be done

from any location by using the Internet as a medium to find information in

libraries around the world that make their data accessible directly to users for free

and anytime

2.2.7 Research is seen from the presence of variables

Research seen from the presence of variables can be categorized in research whose

objects are past, present and period variables that will come. Research whose

objects are past and current variables is also called descriptive research or describe

the variables being studied. Being research object variables that will come, then the

variables are yet but accidentally created by researchers to provide treatment. This

type of research is also called experimental research whose purpose is used to find

causal relationships between the variables studied.

Quantitative, Qualitative and Mixed Method Research Methodology

11

CHAPTER 3

RESEARCH PROPOSAL

3.1 Definition

The research proposal is a means for researchers to communicate their thoughts

about the problems to be studied and serve to convince readers or assessors that

the researcher’s thinking is feasible and will at least provide benefits related to the

relevant disciplines. Because the function of the research proposal, the proposal

should be written-oriented research to the reader / assessor / donor funds .

Furthermore, the research proposal is written using persuasive language so that

those who read in addition to understanding the problem will also easily give

approval for the implementation of the proposal. In writing research proposals,

researchers should also use standard language and straightforward. Although the

purpose is persuasive, avoid long-winded and lengthy languages. The most

important thing is that what we write can be the most effective means of

communicating our ideas so that the reader feels the need to approve them.

3.2 Systematics of Research Proposal

There are various systematic versions of research commonly used by researchers.

The various versions depend on the institution that issued it. Even so there is a red

thread between various versions, including the main things that must be present in

a study, namely the title, formulation of the problem, the purpose of the study, the

methodology used, personnel who conduct, time and cost of research. If it is used

in making research in the student environment, usually for the part of the fee is

eliminated. The following will describe the research proposals issued by the

Directorate of Research Development and Community Service. The author chose

this based on consideration of the scope of the problem, a systematic, easy to

understand sequence and general usage range. The systematics are as follows:

a. Research Title: In general, the research title reflects at least the

relationship between two or more variables. In title writing, it should be

made as short as possible by using straightforward and specific language

so that the reader can easily understand what the researcher will do.

b. Knowledge field: This section provides an explanation of the fields of

study

c. Preliminary: In the introduction, the researcher usually reveals the main

reason why the person chooses a particular problem to be studied so that

the reader can understand the importance of the problem to be examined

from the scientific side. Also in this section, the researcher may write

down the wishes of the researcher to reveal a symptom / concept /

conjecture that is being considered.

Jonathan Sarwono

12

d. Formulation of the problem: In general, the formulation of the problem

is written in the question sentence and the problem formulation should

reflect the relationship of two or more variables. The researcher should

also mention the hypothesis to be tested as well as the approaches,

methods and techniques in answering the problem to be studied.

e. Literature review: In this section the researcher describes the literature

review that underlies the research that will be carried out taken from the

latest reference sources, for example from books or journals. The things

discussed in this section are relevant theories and the results of previous

similar studies. The aim is to avoid discussion of the same problem or

duplication of other people's research.

f. Research purposes: The purpose of the study contains a description that

answers the formulation of the problem above. Besides that also in this

section the researcher can also describe the purpose to explain, prove or

apply a symptom, concept, guess or make a prototype.

g. Research contributions: Here the researcher explains the contribution or

benefits of the research to be carried out in terms of the development of

science, technology and art, solving problems in development and

institutional development.

h. Research methods: This section describes the method used to answer the

problem in detail which includes the variables under study, the research

design used, data collection techniques, data analysis techniques, how to

interpret and infer the results of the study.

i. Implementation Schedule: The research schedule should be written in

detail starting from preparation, preparation of research instruments, data

collection, processing and analysis of data and research reports.

j. Research Personnel: The name of the researcher and his staff (if any) is

written in this section.

Quantitative, Qualitative and Mixed Method Research Methodology

13

k. Estimated Research Costs: Write down the estimated cost of the study in

detail and refer to the specific format applicable in determining the

amount of points that must be paid.

Jonathan Sarwono

14

CHAPTER 4

THE STAGES OF RESEARCH PROCESS

In this chapter, we will discuss the stages of the research process which includes

the following activities:

1. Identifying Problems

2. Make a hypothesis

3. Study of literature

4. Identify and Name Variables

5. Making Operational Definitions

6. Manipulating and Controlling Variables

7. Compile Research Design

8. Identifying and Compiling Observation and Measurement Tools

9. Making a Questionnaire

10. Perform Statistical Analysis

11. Using a Computer for Data Analysis

12. Writing Research Report

4.1 Identifying Problems

What is meant by identifying a problem is that the researcher carries out the first

stage in conducting research, namely formulating the problem to be examined.

This stage is the most important stage in research, because all the research paths

will be guided by the formulation of the problem. Without clear problem

formulation, the researcher will lose direction in conducting research..

4.2 Make a hypothesis

The hypothesis is a temporary answer to the problem that we examine. The

formulation of hypotheses is usually divided into three stages: first, determine the

research hypothesis based on the author's assumptions about the relationship of

the variables being studied. Second, determine the operational hypothesis

consisting of Hypothesis 0 (H0) and Hypothesis 1 (H1). Third, determine the

statistical hypothesis. H0 is neutral and H1 is not neutral. Please note that not all

studies require hypotheses, such as descriptive research.

4.3 Literature Study

At this stage the researcher does what is called a literature review, which is to study

reference books and the results of previous similar studies that have been done by

other people. The aim is to get a theoretical basis for the problem to be studied.

Theory is the basis for researchers to understand the problems that are examined

correctly and in accordance with the scientific framework.

Quantitative, Qualitative and Mixed Method Research Methodology

15

4.4 Identifying and Giving Variable Names

Identifying and naming variables is one of the important stages because only by

knowing the variables being studied, a researcher can understand the relationship

and meaning of the variables being studied.

4.5 Making Operational Definitions

Operational definitions are definitions that make the variables being studied

become operational in relation with the process of measuring these variables.

Operational definitions make it possible for an abstract concept to be an

operational one, making it easier for researchers to take measurements.

4.6 Manipulating and Controlling Variables

What is meant by manipulating variables is to give a treatment to the independent

variable with the aim of the researcher being able to see the effect on the

dependent variable or the variables it influences. While what is meant by

controlling the variable is to control certain variables in the study so that the

variable does not interfere with the relationship between independent variables and

dependent variables..

4.7 Compiling Research Design

What is meant by compiling a research design? Research design, especially in

research that uses a quantitative approach is a tool in research where a researcher

depends on determining the success or failure of the research being carried out.

Research design is like a guiding tool for researchers in carrying out the process of

determining data retrieval instruments, determining samples, collection of data and

analysis. Without a good design, the research carried out will not have high validity.

4.8 Identifying and Compiling Observation and Measurement Tools

What is meant by this section is the stage where a researcher must identify what

tools are appropriate for retrieving data in relation to the purpose of his research.

In research that uses a quantitative approach researchers usually use

questionnaires, especially in Ex Post Facto types of research.

4.9 Making a Questionnaire

In research that uses a quantitative approach, the questionnaire is one of the

important tools for data collection; therefore, researchers must be able to make

questionnaires properly. How to make a questionnaire can be divided into two,

namely in terms of formatting questions and answer models.

4.10 Perform Statistical Analysis

Jonathan Sarwono

16

One prominent feature in research using quantitative approaches is the presence of

statistical analysis. Statistical analysis is used to help researchers know the meaning

of relationships between variables. Until now, statistical analysis is the only tool

that can be scientifically accounted for to calculate the magnitude of the

relationship between variables, to predict the effect of independent variables on

dependent variables, to see the percentage size or the average magnitude of a

variable that we measure..

4.11 Using Computers for Data Analysis

With the development of increasingly sophisticated computer technology and

demanded to conduct research more quickly and the possibility of large amounts

of data, then a researcher needs the help of a computer to do data analysis. A lot of

software has been developed to assist researchers in analyzing data, both data

processing and analysis. One popular program is the IBM SPSS program.

4.12 Writing Report on Research Results

The last stage in research is to make a report on the results of the study in writing.

A written report can be made so that researchers can communicate the results of

research to reader or raised funds.

Quantitative, Qualitative and Mixed Method Research Methodology

17

CHAPTER 5

CHOOSING PROBLEMS AND MAKING

HYPOTHESES

5.1 Introduction

Choosing a problem to study is an important stage in conducting research, because

in essence the entire research process that is carried out is to answer a

predetermined question. Choosing a problem is also not easy because there is no

standard guide . Even so with practice and scientific sensitivity, choosing the right

problem can be done.

How researchers search for problems to be studied, some of the basic guidelines

below will make it easier for us to find problems:

a. Problems should form at least the relationship between two or more

variables

b. The problem must be stated clearly and not double and generally

formulated in the form of question sentences .

c. Problems must be tested using empirical methods, which are possible

data collection that will be used as material to answer the problem being

studied.

d. Problems may not represent moral and ethical position problems.

5.2 Variable Inter- Relationships

Problems should reflect the relationship of two or more variables, because in

practice researchers will examine the effect of one particular variable on other

variables. For example, a researcher wants to know whether or not the influence of

"leadership style" (variable one) on "employee performance" (variable two).

If a researcher uses only one variable in formulating the problem, then the person

concerned only does a descriptive study, for example "leadership style in company

X". Researchers in this case will only conduct studies of existing leadership styles

without considering other factors that influence or are influenced by the leadership

style.

Example: Relationship between employee motivation and work performance

Jonathan Sarwono

18

Motivation: variable one; Job performance: variable two

5.3 Problems Are Formulated Obviously

The problem must be clearly formulated and not double or allow for more than

one interpretation and formulated in the form of question sentences. Examples:

a. Is there a relationship between promotion and sales volume?

b. Does the color of Suzuki motorcycles affect consumers' buying interest?

c. Does mobile phone product design influence consumer buying decisions?

d. Is there a connection between reading interest and the high achievement

index?

The examples above reflect the formulation of a problem that is clear and not

meaningful. In the example "a" the researcher wants to examine the relationship of

promotional variables with the variable sales volume. In the example "b" the

researcher wants to do a study of the relationship of the variable "Suzuki

motorcycle color" to the variable "buying interest". In the example "c" the

researcher will examine the relationship between variables "mobile product design"

with the variable "buying decision". In the example "d" the researcher will examine

the relationship between "reading interest" and the "achievement index".

5.4 Can Be Tested Empirically

The problem must be empirically tested, meaning that the formulation of the

problem made allows researchers to look for data as a means of proof. The main

purpose of data collection is to prove that the problem being studied can be

answered if the researcher searches and collects data. In words, the problem is the

results of the research, the researcher collects the problem.

5.5 Avoid Morals and Ethical Assessments

We recommend that researchers avoid problems related to idealism or values,

because the problem is more difficult to measure compared to problems related to

attitude or performance. For example we will have difficulty measuring problems

such as the following:

Should all students not cheat on the exam? Should all students be diligent in

learning?

It would be better if the problem was made in a form like:

Quantitative, Qualitative and Mixed Method Research Methodology

19

The relationship between test readiness and the value achieved

Effect of student craft on the speed of graduation

5.6 Strategies for Determining Problems

One way to make a good problem formulation is by narrowing the problem from

the very general to the more specific and eventually becoming a specific problem

and ready to be studied.

Sample case:

5.6.1 Recognizing a symptom

The emergence of a sense of dissatisfaction among computer programmers in a

particular company. The company's earnings have continued to increase well over

the past five years. Oral complaints have been received from employees regarding

the payroll structure which is considered to be inadequate.

5.6.2 Identification of problems

a) Evaluating internal and external data by carrying out activities as follows:

Monitoring the dissatisfaction and disseminating information on

company earningsTrack whether there has ever been a sense of

dissatisfaction in the past. Search for literature / references that address

issues similar to those experienced by the company with problems in

other companies.

b) Conduct isolation of the problem area Management does not have a

consistent payroll allocation plan. Based on the interview outside, it is

known that there is dissatisfaction with the payroll system. The directors

have inventoryed complaints from employees regarding payroll

discrimination.

c)

5.6.3 Formulation of the problem

The problem statement will read as follows:

What are the main factors related to payroll levels for professional computer

experts in the company.

Is there a connection between increasing company earnings and dissatisfaction

among programmers.

5.7 Special Considerations in Choosing Problems to Be Researched

In making a selection of problems you can consider the following:

a. Can Be Implemented: If we choose a particular problem, then the

questions below are useful for us to check whether we can or do not do

research with the problem we set: 1) Is the problem within our reach? 2)

Jonathan Sarwono

20

Do we have enough time to do research with this problem? 3) Will we get

access to the sample we will use as respondents as a means of obtaining

data and information? 4) Do we have special reasons so that we believe

we can get answers to the problems we formulate? 5) Have we mastered

the required methods?

b. Research Range: Is the problem sufficient to be investigated? Is the

number of variables enough? Is the amount of data sufficient to be

reported in writing?

c. Linkages: Are we interested in the problem and how to solve it? Are the

problems we are researching related to our knowledge or work

background? If we do research with this problem will we get added value

for our self-development?

d. Theoretical Value: Will the problems that will be examined reduce the

existing theoretical gaps? Will other parties, such as readers or funders,

recognize the importance of this study? Will the results of his research

contribute to the knowledge of the science we are studying? Are the

results of the research worthy of publication?

e. Practical Value: Will the results of his research be practical values for

practitioners in the fields that are suitable for the problem to be studied?

5.8 Making Hypotheses

5.8.1 Definition

After the problem is formulated, the next step is to formulate a hypothesis. What

is the hypothesis? There are many definitions of hypotheses which essentially refer

to the same understanding. Among them is a hypothesis is a temporary answer to

the problem being studied. A hypothesis is a statement that can be empirically

tested, i.e. translation of theory into a testable statement. (Shone, 2022). While

ccording to. Nasution the definition of a hypothesis is "a tentative statement which

is an estimate of what we are observing in an attempt to understand it". (Nasution:

2000)

5.8.2 Function of the Hypothesis

The hypothesis can be derived from the theory relating to the problem that we will

examine. For example, a researcher will conduct a research on the price of a

product so that in order to reduce a good hypothesis, it is better for those

concerned to read the theory regarding price determination.

Quantitative, Qualitative and Mixed Method Research Methodology

21

The hypothesis is a transient truth that needs to be tested for validity, therefore the

hypothesis functions as a possibility to test the truth of a theory. If the hypothesis

has been tested and proved its truth, then the hypothesis becomes a theory. So a

hypothesis is derived from a theory that already exists, then tested the truth and

finally raises a new theory.

The function of the hypothesis according to according to. Nasution is as follows:

1) to test the truth of a theory, 2) provide new ideas to develop a theory and 3)

broaden the knowledge of researchers about a phenomenon being studied.

5.8.3 Considerations in Formulating a Hypothesis

In formulating the hypothesis the researcher needs considerations including:

It must express the relationship between two or more variables, meaning that in

formulating a hypothesis a researcher must have at least two variables to be

studied. Both of these variables are independent variables and dependent variables.

If the variable is more than two, then usually one variable depends on two

independent variables.

It must be stated clearly and not double, meaning that the formulation of the

hypothesis must be specific and referring to one meaning must not cause

interpretation of more than one meaning. If the hypothesis is formulated in

general, then the hypothesis cannot be tested empirically.

It must be able to be empirically tested, the intention is to allow it to be expressed

in an operational form that can be evaluated based on data obtained empirically.

Hypotheses should not reflect moral elements, values or attitudes.

5.8.4 Types of Hypotheses

Broadly speaking there are two types of hypotheses based on the level of

abstraction and form.

According to the level of abstraction the hypothesis is divided into:

a. The hypothesis states that there are similarities in the empirical world:

this type of hypothesis relates to general statements whose truths are

recognized by many people in general, for example "the Javanese people

have a refined attitude and are gentle", "if there is the sound of a dead

animal the dry season begins to arrive, "if it rains the city of Jakarta

Flood". General truths like the one above that have been known by many

people in general, if tested scientifically are not necessarily true.

Jonathan Sarwono

22

b. The hypothesis concerning the ideal model: in reality the world is very

complex, so to study the complexity of the world we need the help of

philosophy, methods, types that exist. Knowledge of authoritarianism will

help us understand, for example in the world of leadership, father's

relationship in educating his child. Knowledge of the idea of nativism will

help us understand the emergence of a leader.

c. The hypothesis is used to find relationships between variables: this

hypothesis formulates the relationship between two or more variables

studied. In compiling the hypothesis, the researcher must be able to know

which variables influence other variables so that the variable changes.

According to the form, the hypothesis is divided into three:

Research / work hypothesis: the research hypothesis is the basic assumption of the

researcher on a problem being studied. In this hypothesis the researcher correctly

considers the hypothesis which will then be empirically proven through hypothesis

testing by using the data obtained during the research. For example: There is a

relationship between the economic crisis and the number of people stressed

Operational hypothesis: the operational hypothesis is an objective hypothesis. This

means that researchers formulate hypotheses not solely based on their basic

assumptions, but also based on their objectivity, that the research hypothesis made

is not necessarily correct after being tested using existing data. Therefore

researchers need a comparison hypothesis that is objective and neutral or

technically called the null hypothesis (H0). H0 is used to give balance to the

research hypothesis because the researcher believes in testing the hypothesis's right

or wrong research depends on the evidence obtained during conduct research.

Example:

H0: There is no relationship between the economic crisis and the number of

people stressed.

Statistical hypothesis: The statistical hypothesis is a type of hypothesis that is

formulated in the form of statistical notation. This hypothesis is formulated based

on researchers' observations of the population in the form of numbers

(quantitative).

For example: H0: r = 0; or H0: p = 0

Quantitative, Qualitative and Mixed Method Research Methodology

23

5.8.5 How to Formulate a Hypothesis

The way to form a hypothesis is with the following stages: formulate the research

hypothesis, operational hypothesis, and statistical hypothesis. The research

hypothesis is a hypothesis that we make and expressed in sentence form.

Example 1: Associative hypothesis

Formulation of the problem:

Is there a relationship between leadership style and employee performance?

Research hypothesis:

There is a relationship between leadership style and employee performance

The operational hypothesis is to define the hypothesis operationally the variables

in it so that it can be operationalized. For example "leadership style"

operationalized- it is a way of giving instructions to subordinates. Employee

performance is operationalized as a high and low income company. The

operational hypothesis is made into two, namely hypothesis 0 which is neutral and

hypothesis 1 which is not neutral

Then the statement of the operational hypothesis:

H0: There is no relationship between how to give instructions to subordinates with

high - low income of companies H1: There is a relationship between how to give

instructions to subordinates with high - low income of the company

The statistical hypothesis is an operational hypothesis that is translated into

statistical numbers according to the measurement tool chosen by the researcher. In

this example the assumption is an increase in income of 30%, then the hypothesis

reads as follows:

H0: P = 0 , 3

H1: P ≠ 0 , 3

Example 2: Descriptive hypothesis

Formulation of the problem: How much is the mastery of English among

students?

Research hypothesis: Mastery of English among students is less than standard

The operational hypothesis:

H0 = Mastery of English among students is the same as standard

H1 = Mastery of English among students is not the same as standard

Statistical hypothesis

H0: r = 80% (0.8)

H1: r ¹ 80% (0.8)

Jonathan Sarwono

24

The standard is assumed to be equal to 80% of the mastery of English.

Example 3: Comparative hypothesis

Formulation of the problem: How is the attitude of students in Bandung to drug

abuse compared to student attitudes in Yogyakarta

Research hypothesis: There are differences in attitudes towards drug abuse among

students in Bandung and students in Yogyakarta

Operational hypothesis:

H0 = There is no difference in the percentage of attitudes towards drug abuse

among students in Bandung and students in Yogyakarta

H1 = There are differences in the percentage of attitudes towards drug abuse

among students in Bandung and students in Yogyakarta

Statistical Hypothesis:

H0: r Bandung = r Yogyakarta

H1 :: r r ¹ Bandung Yogyakarta

5.9 Hypothesis testing

The hypothesis that has been formulated then must be tested. This test will prove

that H0 or H1 will be accepted. If H1 is accepted then H0 is rejected, meaning that

there is a relationship between how to give instructions to subordinates with high -

low income of the company.

Quantitative, Qualitative and Mixed Method Research Methodology

25

CHAPTER 6

STUDY OF LITERATURE

6.1 Aim

The main purpose of conducting a literature study is 1) Finding the variables to be

studied, 2) Distinguishing the things that have been done and determining the

things that need to be done, 3) Synthesizing and obtaining new perspectives, 4)

Determining the meaning and relationship between variable.

The first objective of conducting a literature study is to find the variables to be

studied. In practice, researchers often find it difficult to formulate problems that

deserve to be studied. The problem examined is essentially the variables that will

be examined. Besides helping to identify the problems to be studied, literature

studies can also help researchers in defining variables both conceptually and

operationally and more importantly is helping to identify the existence of

relationships between variables that are conceptually or operationally important to

study.

The second goal is to distinguish between things that have been done and

determine the things that need to be done so that there is no duplication of

research or work in the past that has been done by someone else. Also note that

past research can be material or at least give ideas or inspiration to the research

that will be carried out at this time, especially previous discoveries can give us

direction in conducting current research. We often get a lot of results from past

research suggesting that further / more in-depth research be conducted on the

topics studied.

The third objective is to synthesize and obtain a new perspective, meaning that if a

researcher can carefully synthesize the results of similar studies in the past, it is

possible that the researcher will find something important about the symptoms

being questioned and the ways in which they can be applied in current research

context. In general, researchers prefer things that are specific rather than things

that are general.

The fourth objective is to determine the meaning and relationship between

variables, because all variables studied must be named , defined and put together

with problems that have been formulated along with the hypothesis. If someone

performs the process of defining variables without conducting a library study first,

then it is likely that will diperoleh ialah kesalahan dalam pendefinisian variabel. By

conducting a library study the author in question will get theoretical guidance in

ways to define a variable and also the possibilities of a variable conceptually

defined by previous researchers. Especially in the social sciences and psychology,

in general symptoms or variables are conceptually and operationally defined in

existing theoretical books.

Jonathan Sarwono

26

6.2 Sources

Some library sources that can be used by researchers include 1) abstracts of

research results, 2) indices, 3) reviews , 4) journals 5) reference books. Abstract

research results are valuable reference sources because in abstract researchers

usually write the essence of the research which includes: methods used, problem

formulation, research results and conclusions. By reading the research abstract we

will get an overall picture of the research that has been done. The main advantage

of abstract reading is that we can learn the methods used by these researchers, so

that it inspires us to use similar methods in different contexts and settings.

The index provides book titles compiled based on the main description of each

book but does not provide abstracts, for example the Internet Index will be

displayed as follows: headin section g (news head) Internet, proxy server .

Headings give us information about books on the Internet, the main thing

discussed is about proxy servers.

The review contains writings that synthesize works or books that have been

written in a certain period of time. Posts are arranged by topic and content. In a

review, the author usually provides comparisons and even criticisms of books or

works reviewed by those concerned. Sometimes the author of the review also

provides alternative conclusions to the reader whose purpose is so that the reader

can get a different view from the book he reads.

Journals contain writings in one field of the same discipline , for example

management science in economics or informatics engineering in computer science.

The main use of the journal is that it can be used as a secondary data source

because in general the writings in the journal are the results of research. We can

also use the writing in the journal as a citation material for reference in our

research as well as reference books.

Reference books contain writings that are common in certain disciplines. It's good

we choose books that are reference books that are as a guide in using or making

things. A good reference book will contain deep writing on certain topics and

accompanied by supporting theories so that we will be able to know the

development of theories in the sciences discussed in the book.

6.3 How to search

How to search the library can be done manually or online. If done manually, the

researcher must visit libraries, information sources, such as the Central Bureau of

Quantitative, Qualitative and Mixed Method Research Methodology

27

Statistics. If done online, the researcher must have a computer connected to the

Internet. Due to the absence of censorship on the Internet, we need to evaluate

the quality of writing / books / references that are on the Internet. The following

are discussed ways to measure the quality of writing on the Internet:

a. Reliability: the references sought should be considered reliability,

especially from the side of the author. If the author is indeed an expert in

his field, then the writing can be trusted in its quality. In the back cover of

the book, usually a short history of the author is written, for example the

experience of writing a book, his studies, and his career path. From this

information we can assess how much reliability the book is currently

written in relation to the field of knowledge and experience in writing a

book.

b. CARS (Credibility Accuracy Reasonableness and Support) checklist: Cars checklist

(Robert Harris, 1997) can be used to test the quality of information

originating from the Internet.

First, credibility involves clear and accountable sources of information that allow

us to believe it; clarity of the author's background regarding education, address,

experience, position, and judgment of fellow writers; quality control from fellow

writers; clear references taken from journals or other research results. Second,

accuracy includes not up-to-date, factual, detailed, definite, comprehensive, reader-

oriented and purpose-oriented, making current sources not expired information,

and can provide a full picture of the truth. Third, it can be accepted with common

sense which includes fair and impartial, provides balance, is objective, does not

give rise to conflicts of interest, is not incitement; has the purpose to be used as a

source that can be trusted because it raises the whole truth. Fourth, the existence

of support such as reference sources, contact information, enabling service

demands, the aim is to provide convincing evidence to the reader if the reader

makes a claim.

Jonathan Sarwono

28

CHAPTER 7

IDENTIFYING AND GIVING VARIABLE

NAMES

7.1 Definition

Variables are defined as "something that may vary or differ" (Brown, 1998: 7).

Another definition that is more detailed says that the variable "is simply symbolic

that can assume any one of a set of values" (Davis, 1998: 23). While Shone (2022)

define a variable as something that varies over time or over subjects (in other

words, varies within the study), also used to mean the operational definition of a

concept (how do we measure something).

The first definition states that variables are something different or varied,

emphasizing the word something is clarified in the second definition, namely a

symbol or concept which is assumed to be a set of values. The abstract definition

will be clearer if given an example as follows:

a. Relations between intelligence and learning achievement

b. Influence of colors on interest in buying a motorcycle

c. Relationship between promotion and sales volume

Examples of variables are: intelligence, learning achievement, color, buying

interest, promotion and sales volume

7.2 Variable Types

7.2.1 Independent Variables: Independent variable is a stimulus variable

orvariable that affects other variables. Independent variables are variables

that are the variables are measured, manipulated, or chosen by the

researcher to determine the relationship with a symptom observed. In the

example above, "color" is an independent variable that can be manipulated

and seen its influence on "buying interest", for example whether the red

color of a motorcycle can lead to consumer buying interest in the

motorcycle.

7.2.2 Dependent Variables: Dependent variable is a variable that gives a

reaction / response if it is associated with a free variable. Dependent

variable is a variable whose variables are observed and measured to

determine the effect caused by the independent variable. In the example of

the influence of color on motorbike buying interest, the dependent

variable is "buying interest". How big is the influence of red on consumer

buying interest in the motorcycle. To ensure the influence of red

Quantitative, Qualitative and Mixed Method Research Methodology

29

independent variables on buying interest, the red color can be replaced

with blue. If the magnitude of the effect is different, the manipulation of

the free variable proves that there is a relationship between the color

independent variable and consumer buying interest.

7.2.3 Relationship Between Independent Variables and Dependent

Variables

In general, people do research using more than one variable, namely independent

variables and dependent variables. The two variables are then searched for

relationships

Example 1

Research hypothesis: There is a relationship between "leadership style" and

"performance" employees

Independent variable: leadership style

Dependent variables on buying interest

Leadership style has a relationship with employee performance, for

example a centralized leadership style will have an impact on employee

performance is different from the delegative leadership style.

Example 2

Research hypothesis: There is a relationship between promotion and sales

volume

Independent variable: promotion

Dependent variable on: sales volume

Promotion has a relationship with the presence and absence of an increase

in sales volume in certain companies.

7.2.4 Moderate Variables

The moderate variable is the second independent variable that is deliberately

chosen by the researcher to determine whether his presence influences the

relationship between the first independent variable and the dependent variable.

Moderate variables are variables whose variables are measured, manipulated, or

chosen by the researcher to find out whether the variable changes the relationship

between independent variables and dependent variables.

In the case of a relationship between motorcycle colors and buying interest,

researchers chose the moderate variable as "price". By entering the moderate price

Jonathan Sarwono

30

variable, the researcher wants to know whether the magnitude of the relationship

between the two variables changes. If it changes then the existence of moderate

variables plays a role, while if it does not change the moderate variable does not

affect the relationship between the two variables studied.

Another example:

Hypothesis: There is a relationship between promotion on television media

with the increasing awareness of Samsung mobile phone brands among

consumers

Independent variable: promotion

Dependent variable: brand awareness

Moderate variable: media promotion

7.2.5 Control Variable

In research researchers always try to eliminate or neutralize the influence that can

disrupt the relationship between independent variables and dependent variables. A

variable whose influence will be removed is called a control variable Control

variables are defined as variables whose variables are controlled by researchers to

neutralize their effects. If not controlled, these variables will affect the symptoms

being studied.

Example:

Hypothesis: there is an influence of shirt color contrast on buying decisions among

women

Independent variable: color contrast

Dependent variable: buying decisions

Control variable: gender

In the case of research above the control variable of female sex. Assuming

researchers only women are affected by the color contrast if they want to buy it.

7.2.6 Intervening Variables

Independent, dependent, control and moderate variables are concrete variables.

The three variables, namely the independent, control and moderate variables can

be manipulated by the researcher and the influence of the three variables can be

seen or observed. As with the intervening variable is hypothetical, meaning that

the effect is not concrete, but theoretically can affect the relationship between

independent variables and depending on what is being studied. Therefore, an

intermediate variable is defined as a variable that theoretically influences the

relationship of variables being studied but cannot be seen, measured, and

Quantitative, Qualitative and Mixed Method Research Methodology

31

manipulated; the effect must be inferred from the effects of independent variables

and moderate variables on the symptoms being studied. Fot example:

Hypothesis: If the interest in the task increases, the performance of performing the

task will increase

Independent variable: interest in the task

Dependent variable: the performance of the task

Intervening variable: learning process

The description of the case above is as follows: If students are interested in the

assignments given by the lecturer, the results will be good. The size of the

performance is influenced by interest; however the final results of the assignment

are influenced by the factors students learn or not in doing the task first. With high

interest and good preparation for learning, the performance will be even greater.

7.2.7 Variable Relationship Schemes

The relationship scheme between variables shows the influence of independent,

moderate, control and intermediary variables on dependent variables.

The model scheme was made by Brown (1988) as follows:

ABOUT THE AUTHOR

The Brown scheme can be read as follows: the central relationship in the study is

between independent variables and dependent variables. The arrows show more

the direction of focus of the researchers' thinking and research design, rather than

the causal relationship. Thus focus variable is the dependent variable. In the early

stages of the research it was carried out only to determine the effect of

independent variable on dependent variable. The intervening variable serves as

label for the relationship between the two variables or the process that connects

the independent variables and dependent variable but is not observed. Researchers

may also consider the existence of other independent variable, namely a moderator

Independent

Variabel

Dependent

Variabel

Moderate

Variabel

Control

Variabel

Intervening

Variabel

Jonathan Sarwono

32

variable that will used to determine whether there will be a change in the

relationship between the independent variable and the dependent variable if the

moderator variable is included in the study. The researcher may also control other

independent variables if the person concerned wants to neutralize, or eliminate the

influence of the control variable.

7.2.8 Case Examples

Measuring methods in teaching on student achievement. The assumption of the

researcher is that there are other variables that influence, namely the student's

personality, gender and means of formality in the class.

Independent variable: method

Dependent variable: learning achievement

Moderator variable: student personality

Control variable: gender

Intervening variable: means of formality in class

Information from the above cases are as follows: The researcher wants to know

whether or not there is an influence of the teaching method on student

achievement. The teaching method is an independent variable and student

achievement is a dependent variable. Researchers also consider the existence of

other factors that influence the relationship of these two variables, namely student

personality. Student personality variables are deliberately chosen to determine

whether their presence influences the relationship between independent and

dependent variables. Researcher intends to neutralize the possibility of influencing

sex factors, therefore sex will be controlled as a control variable. The aim is to

eliminate the possibility of confusion due to these factors. In theory the means of

formality in the classroom will influence the relationship between teaching

methods and student achievement. Then the means of formalities in the class are

used as intermediary variables.

7.2.9 Paradigm of Relations Between Variables

The emphasis of research that uses a quantitative approach is the pattern of

relationships between the variables being studied. This understanding is based on

the philosophy of positivism which says that symptoms can be classified and

symptoms have a causal or causal relationship. Therefore, in conducting research,

researchers must be able to understand and find relationships between variables.

Because the symptoms being studied can be seen by looking at the relationships

between variables. The paradigm of relations between variables according to

Sugiyono (2002) will be discussed further in this section, the author uses examples

that have been adjusted, namely:

Quantitative, Qualitative and Mixed Method Research Methodology

33

a. Simple Paradigm of Two Variable Relationships

Research that uses this paradigm only examines the relationship between one

independent variable and one dependent variable.

Example:

Independent variable (X): promotion

Dependent variable (Y): sales volume

The relationship of these two variables can be described as follows:

The above paradigm will produce research as follows:

1 Problem Formulation: How much is the relationship (or the influence of

advertising on sales volume?)

2 The theory needed by researchers is the theory of advertising and sales

3 The research hypothesis will be as follows: There is a relationship between

advertising and sales volume

4 Data Analysis Techniques: to see the relationship between X and Y can be used

Pearson Product Moment correlation while testing the hypothesis can be used

to test the significance of product moment correlation.

b. Relationship Paradigm More Than Two Sequential Variables

Research that uses this paradigm will examine more than one independent variable

with one dependent variable. The relationship between variables is still simple,

namely sequentially, meaning that the condition of the independent variable 2 is

the result of the existence of independent variables 1. In other words, the

independent variable 1 affects the independent variable 2; independent variables 1

and 2 affect dependent variables.

Example:

Independent variable 1 (X1): Quality of wired network

Independent variable 2 (X2): Service quality

Dependent variable (Y): Customer satisfaction

The relationship pattern of these variables can be described as follows:

The relationship pattern can be explained as follows: the quality of the cable

network affects the quality of service. The quality of the cable network and service

X

Y

X1

Y

X1

Jonathan Sarwono

34

quality affect customer satisfaction.

c. Relationship Paradigm Between Two Independent Variables with One

Dependent Variable

Research that uses this paradigm will examine the relationship of two independent

variables with one dependent variable.

Example:

Independent variable 1 (X1): IQ

Independent variable 2 (X2): Motivation

Dependent variable (Y): Test results

The relationship pattern of these variables can be described as follows:

The pattern of relationships between these variables can be explained as follows:

IQ affects test results partially

Motivation affects the results of the exam partially

IQ and motivation affect test results simultaneously

d. Relationship Paradigm Three Independent Variables with One Dependent

Variable

Research that uses this paradigm will examine the relationship of three

independent variables with one dependent variable.

Example:

Independent variable 1 (X1): Salary

Independent variable 2 (X2): Career path

Independent variable 3 (X3): Employee recruitment system

Dependent variable (Y): Work performance

The relationship pattern of these variables can be described as follows:

X1

X2

Y

X1

X2

Y

X3

Quantitative, Qualitative and Mixed Method Research Methodology

35

The pattern of relationships between these variables can be explained as

follows:

Salary affects work performance partially

Career paths affect work performance partially

Employee recruitment system influences work performance partially

Salaries, career paths and employee recruitment systems affect work

performance simultaneously

e. Paradigm Relationship One Independent Variable with Two Dependent

Variables

Research that uses this paradigm will examine the relationship of one independent

variable with two dependent variables.

Example:

Independent variable (X): Education level

Dependent variable (Y1): Work achieved

Dependent variable 2 (Y2): Insight

The relationship pattern of these variables can be described as follows:

The pattern of relationships between these variables can be explained as follows

The level of education affects the work achieved and insight

f. Relationship Paradigm Between Two Independent Variables with Two

Dependent Variables

Research that uses this paradigm will examine the relationship of two independent

variables with two dependent variables.

X

Y1

Y2

Jonathan Sarwono

36

Example:

 Independent variable (X1): Service speed

 Independent variable (X2): Product price

 Dependent variable 1 (Y1): Number of customers

 Dependent variable 2 (Y2): Decision to buy

The relationship pattern of these variables can be described as follows:

The pattern of relationships between these variables can be explained as follows:

 Service speed affects the number of customers

 Service speed affects the decision to buy

 Product prices affect the number of customers

 Product prices affect buying decisions

g. Path Paradigm

Research that uses this paradigm will examine the relationship of three

independent variables, one of the independent variables functions as an

"intervening variable" with one dependent variable. The possibility of the influence

of X1 and X2 on Y can be directly, but also the another possibility is that X1 and

X2 affect Y after going through X3.

Example:

Independent variable (X1): IQ

Independent variable (X2): Learning pattern

Independent variable (X3): Motivation

Dependent variable (Y): Learning achievement

The relationship pattern of these variables can be described as follows:

Quantitative, Qualitative and Mixed Method Research Methodology

37

The pattern of relationships between these variables can be explained as follows:

IQ affects learning achievement

Learning patterns affect learning achievement

Motivation affects learning achievement

IQ and learning patterns with intermediary motivation affect learning

achievement.

CHAPTER 8

DEVELOPING OPERATIONAL VARIABLE

DEFINITIONS

8.1 Importance of Variable Operationalization

Variables must be operationally defined so that the relationship between one

variable and another is easier to find and its measurement. Without the

operationalization of variables, researchers will experience difficulties in

Jonathan Sarwono

38

determining the measurement of the relationship between variables that are still

conceptual.

Operationalizing variables is useful for: 1) identifying observable criteria that are

being defined; 2) shows that a concept or object might have more than one

operational definition; 3) knowing that operational definitions are unique in

situations where the definition must be used.

8.2 Operational definitions are based on observable criteria

What is meant by operational definition is a definition based on observable

characteristics of what is being defined or "changing concepts in the form of

constructs with words that describe behavior or symptoms that can be observed

and which can be tested and determined by people others "(Young, quoted by

Koentjaraningrat, 1991; 23). The emphasis of the definition of operational

definition is on the word "can observed ". If a researcher makes an observation of

a symptom or object, other researchers can also do the same thing, namely

identifying what has been defined by the first researcher.

Unlike the conceptual definition, the conceptual definition is more hypothetical

and "not observable". Because conceptual definition is a concept defined by

reference to another concept. Conceptual definitions are useful for making the

logic of the hypothesis formulation process.

8.3 Ways to Arrange Operational Definitions

There are three approaches to compiling an operational definition, namely called

Type A, Type B and Type C.

8.3.1 Type A Operational Definition

Type A operational definitions can be arranged based on the operations that must

be performed, so that the symptoms or conditions that are defined become real or

can occur. Using certain procedures researchers can make symptoms become

apparent. For example: "conflict" is defined as a situation produced by placing

two or more people in a situation where each person has the same goal, but only

one person will be able to achieve it.

8.3.2 Operational Definition of Type B

Type B operational definitions can be compiled based on how certain objects that

are defined can be operationalized, namely in the form of what they do or what

composes their dynamic characteristics. For example: "smart people" can be

defined as someone who gets high values in his school.

8.3.3 Type C Operational Definition

Type C operational definitions can be arranged based on the appearance of what

the object or symptom is defined, namely what constitutes the static characteristics.

For example: "Smart people" can be defined as people who have strong memories,

mastered several foreign languages, have good thinking skills, are systematic and

have the ability to calculate quickly.

Quantitative, Qualitative and Mixed Method Research Methodology

39

8.4. Criteria of Uniqueness

In compiling operational definition, the definition should be able to identify a set

of unique criteria that can be observed. The more unique operational definition is,

the more useful it will be. Because of the definition of information to researchers,

and more often than not, they will not be included in the definition of accidentally

and can increase. Even so, the uniqueness / specificity does not become a barrier

to its general applicability.

CHAPTER 9

TECHNIQUES FOR MANIPULATING AND

CONTROLLING VARIABLES

One characteristic of experimental research is that it allows researchers to

manipulate and control variables that cannot be carried out in the types of

descriptive or exploratory research. Variable manipulation means that the

researcher gives a certain treatment to the independent variables which will have

Jonathan Sarwono

40

an effect on the dependent variable. The purpose of manipulating an independent

variable is that the researcher wants to see how much the effect of giving different

treatment of independent variables on the dependent variable affected. By

manipulating independent variables, the researcher will can know which treatment

is the most effective result. Controlling variables means that researchers exercise

control in such a way that researchers can eliminate the influence of these variables

so as not to influence the process of measuring the influence of the variables under

study. The purpose of controlling variables is to eliminate the bias that is likely to

arise because the influence of these variables is not desired by the researcher.

Below will be discussed techniques for manipulating and controlling variables.

9.1 Control Group

One technique in manipulating and controlling variables is by way of making a

control group (or comparator ) with the existence of a controlling group, the

researcher will be able to control the possibility of the emergence of factors that

can influence the valid assessment process on the effects of the treatment

conditions imposed on the group or object being studied.

9.2 Factors Affecting Internal Validity

In experimental research, if not controlled, it will affect what is called internal

validity. External variables that can affect internal validity will be discussed below,

including:

9.2.1 History : Factor of history refers to events that are happening in the

environment at the same time when the variables being made in the

experiment are being tested or carried out measurements. This can be

understood by using an example as follows: if a member of a group

was having a psychological problem or external pressure to include in

the study such as testing a new curriculum on the group, the results of

test measurements experimentally may not be reflecting the problems

of testing the implementation of the new curriculum but reflect

external event factors or called " external historical event "

9.2.2 Selection : The selection process that is not good will result in a group

that is being tested there is a difference in the ability to accept,

respond, age, type of work and so on. The result is a different response

to the treatment being tested. The process of selecting group members

who are investigated incorrectly will produce wrong or biased

conclusions.

9.2.3 Maturation: Maturation means that there is a process of change that

Quantitative, Qualitative and Mixed Method Research Methodology

41

occurs in the object being studied (respondent) at their time is

participating in experimental research. Usually this happens in research

that requires a long time. People who are the object of research or

respondents constantly change both physically and mentally. Changes

that occur in the respondent can result in bias in the measurement

results.

9.2.4 Testing: Testing refers to the effects that occur because of a pre-test

that precedes the actual test that will be imposed on the respondents.

This pre-test activity will affect the respondents in working on the

actual test. There is a possibility that there is a tendency for individuals

who have pre-tested the results to be better at working on the actual

test.

9.2.5 Instrumentation: Instrumentation has the understanding that changes

that occur in measurement or observation procedures during the

experiment take place. These procedures include tests, mechanical

measurement instruments, officers who make observations or those

who make judgments. One form of threat that can affect internal

validity is for example: if the observation officer, the appraiser or

interviewer either intentionally or unintentionally matches the desired

research hypothesis. This event allows the officers to direct the

respondent in the desired hypothesis.

9.2.6 Experimental Mortality: In a study that requires post-test data, all

respondents were included in the study because respondents who

resigned from the experimental study differed from respondents who

were still active in the group; hence this status difference will give rise

to what is called post-test bias (post-test bias) or bias because there is

no internal validity based on mortality. For example, in an experimental

study, researchers used graduates from two different disciplinary

groups, for some reason some members of each group were absent or

retreated so that one of the groups lost more members than the other

group. Because the number of the two groups is not the same, it will

cause bias.

9.2.7 Stability: What is meant by stability bias is that the research findings

cannot be scientifically unreliable. This bias can be tested using

statistical calculations.

Jonathan Sarwono

42

9.2.8 Interactive Combinations of Several Factors: There is also the

possibility of the results of research where there is no internal validity

caused by several combination factors simultaneously, for example the

source loses internal validity due to the interaction of the combination

of completion factors and maturation. In determining group selection

based on age, some members are mature while other members are still

at the stage of physical or mental change.

9.2.9 Hope: What is meant by the expectation factor is the expectation of

the researcher to obtain certain research results according to his

wishes. In order for the purpose to be achieved, the researcher

consciously or unconsciously tries to influence the research process

and the object being studied. Such actions can result in a loss of

internal validity factors.

9.3 Factors Affecting External Validity: External validity means the existence of

a generality or ability to represent (population) the results of research. The results

of the study can be applied in the context of time, place and group of people

(objects research) that is different. Only studies that have external validity can be

said to reflect the population. Below will be discussed several factors that can

influence the presence or absence of external validity in the current research.

9.3.1 Reactive Impact of Testing : If the researcher wears a pretest activity

that can affect the respondents being studied in an experimental study,

then the impact of the treatment can be influenced by some of the

pretest's activities. If the pretest is not done, the impact of the

treatment will not be the same.

9.3.2 Effects of Selection Bias Interaction : If the researcher makes a

mistake in sampling which results in the sample not representing a

larger population, the researcher will have difficulty in generalizing his

study findings from the sample level to the population. Example: if the

researcher takes a sample from a part of city A, then the result will not

be valid if applied to the other parts of the city.

9.3.3 Reactive Effects of Experimental Settings : Researchers in

conducting experimental arrangements intentionally or unintentionally

can create a contrived condition to limit the possibility of research

results that can be generalized in testing a treatment that is not

experimental.

Quantitative, Qualitative and Mixed Method Research Methodology

43

9.3.4 Multiple Inference Treatment : In conducting the study the

researcher gave several treatments simultaneously to the respondents

where the treatments could be in the form of experimental or non-

experimental treatments; these treatments can interact with various

ways so that it can cause representation of the impact of the treatment

to decrease.

9.4 Equating Experimental Groups and Controllers: To get good experimental

results, researchers need to select the control group members who have similar

characteristics with the experimental group members. By doing the selection

process, the possibility of losing selection validity will decrease. The following will

be given examples of several ways to equalize the experimental group and the

control group: a. Randomization: Randomization technique is a procedure to

control selection variables without identifying them first. The aim is to avoid the

possibility of different types of people being chosen as members in the control

group or the controlling group. b. Matching Pair Techniques:

Before using a suitable pairing technique, researchers must first determine which

control variables can be applied to different individuals. Usually the variables used

are gender, age, socio-economic status, IQ, achievement; while in dependent

variables, researchers usually use pretest values. In determining partner members,

for example researchers make a partner between one member who is 30 years old

with other members who have the same age. All members who are used as

research objects are made pairs. So if the researcher needs 50 members it will be 25

pairs. Then by making a random selection one member of each pair will be made a

member in the experimental group, while the other members will be made

members of the control group. c. Suitable Group Techniques: Using a suitable

group technique can be done in the same way as the matching pair technique. In

the experimental group certain members are selected using the same variable, for

example age; then the control group is selected using the same variable. d. Limiting

Population Research samples are basically taken from the population. By limiting

population characteristics, researchers will automatically be able to limit the

characteristics of the sample. For example, researchers want to do research using

the object of college students, it should be restricted, for example only university

students, then limited to only university students majoring in engineering. This

limitation will produce the same characteristics in the population and if the sample

is drawn from the population, the sample will have the same characteristics.

CHAPTER 10

MAKING RESEARCH DESIGN

1.0.1 Introduction

In conducting research one of the important things is to make a research design.

Research design is like a road map for researchers who guide and determine the

direction of the research process in a correct and appropriate manner in

accordance with the objectives set. Without a correct design, a researcher will not

be able to do a good research because he does not have a clear direction. In order

Jonathan Sarwono

44

to achieve the correct design creation, the researcher needs to avoid potential

sources of errors in the overall research process. These errors are:

a. Error in Planning: Errors in planning can occur when researchers make

mistakes in compiling designs that will be used to gather information.

This error can also occur if the researcher is wrong in formulating the

problem. Errors in formulating problems will produce information that

cannot be used to solve the problem being studied. The way to overcome

this error is to develop a good and correct proposal that clearly specifies

the method and value added of the research that will be carried out.

b. Error in Data Collection: Errors in data collection occur when

researchers make mistakes in the process of collecting data in the field.

This error can increase the level of errors that have occurred due to

improper planning. To avoid this, the collected data must be a

representation of the population being studied and the data collection

method must be able to produce accurate data. The way to deal with this

error is to be careful and accurate in carrying out research designs that

have been designed in the proposal.

c. Error in Analyzing: Errors in conducting an analysis can occur when the

researcher is wrong in choosing how to analyze data. Furthermore, this

error is caused by an error in choosing an analysis technique that is

appropriate to the problem and available data. The way to overcome this

problem is to make a justification of the analytical procedure used to infer

and manipulate data.

d. Error in Reporting: Errors in reporting occur if the researcher makes a

mistake in interpreting the results of the study. Errors like this occur

when giving meaning to the relationships and numbers identified from

the data analysis stage. The way to overcome this error is the results of

data analysis examined by people who are truly experts and master the

problem of the results of the research.

10.2 Types of Research Design

Types of research designs are as follows:

The exploratori research design is used for initial research which functions to

explain and define a problem. Early research is not to seek final conclusions.

Included in this category are surveys conducted by experts, case studies, analysis of

secondary data and research using a qualitative approach.

Quantitative, Qualitative and Mixed Method Research Methodology

45

Conclusive research design is used for descriptive research and ex-experimental

research. Descriptive research serves to describe the characteristics / symptoms /

functions of a population. The methods used are usually surveys and observations.

Descriptive research has the characteristics of a hypothesis, a structured and

inflexible design, prioritizing accuracy and understanding of previous problems.

Causal research is used to identify causal relationships between variables that

function as causes (independent variables) and which variables function as

consequential variables (dependent variables).

Broadly speaking, in a conclusive quantitative flow there are two types of designs,

namely: Ex Post Facto Design and Experimental Design. The factors that

distinguish these two designs are that in the first design there was no manipulation

of the independent variables while in the second design there was manipulation of

the independent variables. The main purpose of using the first design is

exploratory and descriptive; while the second design is explanatory. If viewed in

terms of the level of understanding of the problem under study, the ex post facto

design results in a level of understanding of the problem under study at the surface

level, while experimental designs can produce a deeper level of understanding. The

two main designs have more specific sub-designs. Included in the first category are

field studies and surveys. While what is included in the second category is an

experiment in the field (field experiment) and experiments in the laboratory

(laboratory experiment)

10.2.1 Sub Design Ex post Facto

a. Field Study: Field studies are research designs that combine literature

searches, experiential surveys and / or case studies where researchers

attempt to identify important variables and relationships between these

variables in a particular problem situation. Field studies are generally used

as a means of further and in-depth research.

b. Survey: Survey design depends on the use of the type of questionnaire.

Surveys require a large population if researchers want the results to reflect

real conditions. The larger the sample, the more surveys provide more

accurate results. With a survey a researcher can uncover many problems,

even if only on the surface. Even so, surveys are useful if researchers

want information that is diverse and diverse. Survey methods are very

popular because they are widely used in business research. Another

advantage of the survey is that it is easy to implement and can be done

quickly.

10.2.2 Experimental Sub Design

a. Field Experiments: Field experiment design is a study conducted by using

a realistic setting where researchers intervene and manipulate

independent variables.

b.

c. Laboratory Experiments: The laboratory experimental design uses

Jonathan Sarwono

46

artificial background in conducting his research. By using this design,

researchers intervene and manipulate independent variables and allow

researchers to control the main aspects of errors.

10.3 Validity

Validity is related to the problem of limiting or suppressing errors in research so

that the results obtained are accurate and useful to implement. There are two

validities, namely internal validity and external validity.

a. Internal Validity : Internal validity is the level at which research results

can be trusted truthfully. Internal validity is an essential thing that must

be fulfilled if the researcher wants the results of his study to be

meaningful. In connection with this, there are several things that become

obstacles to obtaining internal validity: a) History: This factor occurs

when external events in the investigation conducted affect the results of

research. b) Maturation: There are changes that occur in the respondent

in a certain period of time, such as increasing age or the existence of

fatigue and saturation factors. c) Testing: Effects produced by the process

under investigation which can change the attitude or actions of the

respondent. d) Instrumentation: Effects that occur due to changes in

tools when conducting research. e)Selection: An artificial effect where the

selection procedure influences the results of the study. f)Mortality: The

effect of the existence, loss or departure of the respondents studied

b. External Validity : External validity is the degree to which research results

can be generalized to populations, settings and other things in similar

conditions. Things that become sources of external validity are: a) Testing

Interactions: Artificial effects made by testing respondents will reduce

generalizations in situations where there is no testing on the respondent.

b) Selection Interaction: The effect in which the types of respondents

that influence study results can limit their generality. c) Interaction

Setting: Artificial effects made using certain backgrounds in the study

cannot be replicated in other situations.

10.4 Ex Post Facto and Experimental Specific Designs

Before discussing Ex Post facto and experimental specific designs, the notation

system used needs to be known first. The notation system is as follows:

X: Used to represent the exposure (exposure) of a group that is tested against

an experimental treatment on an independent variable which then effects

on the dependent variable will be measured .

O: indicates the existence of a measurement or observation of dependent

variables that are being studied in certain individuals, groups or objects.

R: indicates that individuals or groups have been randomly selected and

determined for study purposes.

10.4.1 Ex Post Facto

Quantitative, Qualitative and Mixed Method Research Methodology

47

As mentioned earlier that in the Ex Post Facto design there is no treatment

manipulation of the independent variables then the notation system, whether field

studies or surveys are only written with O or O more than one.

Example 1: The study was conducted using two populations, namely

Company A and Company B, the notation:

O 1

O 2

Where O 1 is an observation activity carried out in company A and O 2 is an

observation activity carried out in company B. For example: We randomly

examined 200 companies from a population of 1000 companies regarding their

payroll system. The survey was conducted by sending questionnaires to 200

managers, so the design configuration would be as below:

(R) O 1

Where O 1 represents a survey in 200 companies by giving questionnaires to 200

managers selected randomly (R).

If we examine the same sample repeatedly, for example three times in three

consecutive months, the notation is:

(R) O 3 where O 1 is the first observation, O 2 is the second observation and O 3

is the third observation.

10.4.2 Experimental Designs

The experimental design was divided into two, namely: pre-experimental (quasi-

experimental) and actual experimental design (true experimental). The difference

between the two types of design lies in the concept of control.

a. One Shot Case Study: The simplest experimental design is called One

Shot Case Study. This design is used to research one group by being

given one treatment and the measurement is done once. The diagram is

as follows:

X O

b. One Group Pre-test - Post-test Design: The second design is called One

Group Pre-test - Post-test Design which is a development of the design

above. The development is by doing one measurement in front (pre-test)

before the treatment (treatment) and after that the measurement is done

again (post-test). The designs are as follows:

O 1 XO 2

In this design the researcher makes initial measurements on an object

under study, then the researcher gives certain treatment. After that the

measurement is done again for the second time. The design can be

developed in other forms, namely: "time series design". If the

Jonathan Sarwono

48

measurement is done repeatedly in a certain period of time. Then the

design will be as below:

O 1 O 2 O 3 XO 4 O 5 O 6

In the time series design, researchers took measurements in front for 3

times in succession, then he gave treatment to the object under study.

Then the researchers took measurements again 3 times after the

treatment was done.

c. Static Group Comparison: The third design is Static Group Comparison

which is a modification of design b. In this design there are two groups

chosen as the object of research. The first group received treatment while

the second group received no treatment. This second group functions as

a comparison / control group. The designs are as follows:

XO 1 O 2

d. Post Test Only Control Group Design: This design is the simplest design

of the actual experimental design (true experimental design), because the

respondents were actually randomly selected and treated and there were

the control group. This design has fulfilled the actual experimental

criteria, namely by the presence of variable manipulation, random

selection of groups studied and treatment selection. The designs are as

follows:

(R) XO 1

(R) O 2

The purpose of the design is that there are two groups chosen randomly.

The first group was treated while the second group was not. The first

group was given treatment by the researcher then carried out

measurements; while the second group used as the control group was not

treated but only measured.

e. Pre-test - Post-test Control Group Design: This design is a development

of the design above. The difference lies in both the first group and the

control group in the pre-test. The designs are as follows:

(R) O 1 XO 2

(R) O 3 O 4

f. Solomon Four Group Design: This design is a combination of Post Test

Only Control Group Design and Pre-test - Post-test Control Group

Design which is an ideal design model for conducting controlled

experimental research. The researcher can press as little as possible the

sources of error because there are four different groups with six

Quantitative, Qualitative and Mixed Method Research Methodology

49

measurement formats. The designs are as follows:

(R) O 1 XO 2

(R) O 3 O 4

(R) XO 5

(R) O 6

The design intent is: The researcher randomly selected four groups. The

first group which is the core group is given treatment and two

measurements, namely in front (pre-test) and after treatment (post-test).

The second group as a control group was not treated, but measurements

were taken as above, namely: measurement in front (pre-test) and

measurement after treatment (post-test). The third group was treated and

only one measurement after treatment (post-test) and the fourth group as

the control group of the third group was only measured once.

10.4.3 Advanced Experimental Design

a. Completely Randomized Design: This design is used to measure the influence of

an independent variable manipulated on dependent variables. Random group

selection is done to get equivalent groups. Case: The directors of a company

want to know the influence of three different types of giving instructions made by

superiors to subordinates. For the purpose of this study three randomly selected

groups of 25 people were selected. Instructions for the first group were given

orally, for the second group in writing and for the third group the instructions

were not specific. The three groups were given about 15 minutes to think about

the situation. Then all three were given objective tests to find out how well they

understood the work to be done.

The formulation of this case problem is: Does the manipulation of independent

variables affect the understanding of subordinate employees in carrying out their

work?

The purpose of this study is to determine which types of instructions can create a

better understanding of the work ordered by superiors.

Research Design:

Instruction

A1. (Verbal)

A2. (Written)

A3. Not Specific

X11

X12

X13

X21

X22

X23

Jonathan Sarwono

50

X31

X32

X33

X25,1

X25,2

X25,3

Treatment

x.1

x.2

x.3

b. Random Block Design: This design is an improvement in Random Perfect

Design above. In the previous design the differences found in each individual were

not considered, resulting in groups that had different characteristics. So that the

design that we make can produce good output, it is necessary to choose group

members (respondents) who come from populations that have the same

characteristics. Therefore researchers must be able to identify some of the main

sources of the intended differences early.

Examples of Design:

Treatment

Experimental Group

Control Group

Instruksi

A1. (Verbal)

A2.

A3. Not

Average

Blok

(Written)

Spesifik

Blok

(Departement)

B1

5 (Workers)

X1.

B2

5 (Workers)

X2.

B3

5 (Workers)

X3.

B4

5 (Workers)

X4.

B5

5 (Workers)

X5.

Average

x.1

x.2

x.3

Treatment

The above design can be explained as follows: At the time the study was

conducted using the previous design, members from three groups came from

different backgrounds. Member background differences are a nuisance or are

referred to as interfering variables. For this reason, it is necessary to equalize the

members of each group. The trick is to create blocks that function to get members

of the same group. In this case the block is determined based on the department

(section) where the group members originate. Furthermore, workers from the

same department are divided into five based on their respective departments. Then

each group received the same treatment , namely the first group received oral

instructions, the second group received written instructions and the third group

the instructions were not specific. By using this design, the researcher will be able

to see the effects caused by the system of blocks per department and the

interaction of instructions for the three groups.

Quantitative, Qualitative and Mixed Method Research Methodology

51

c. Latin Square Design: This design is used to control two confounding variables at

once. In connection with the above case, there is still one other confounding

variable, namely "the ability of workers". The variable capabilities of our workers

divide into three levels, namely: high ability, medium ability and low ability. These

three levels of capability variables are then placed in the rows and columns of the

Latin Square model . This design consists of three rows and three columns. Then

randomly taken 3 employees from each department.

The design is as below:

Ability of Workers

Block

c1 Height

c2 Intermediate

c3. Low

Average

B1

(a1) x1

(a2) x1

(a3) x1

X1 ...

B2

(a2) x2

(a3) x2

(a1) x2

X2 ...

B3

(a3) x3

(a1) x3

(a2) x3

X3 ...

c. Factorial Design: Factorial design is used to evaluate the impact of the

combination of two or more treatments on dependent variables. In the

case below, factorial analysis was applied using a perfect random design

with 3 rows and 3 columns. The research case is as follows: researchers

want to see two independent variables, namely the variable "level of

contrast" and "line length" of an ad. Contrast levels are manipulated to

"low", "medium" and "high"; while the length of the line is manipulated

to "5 inches", "7 inches" and "12 inches". The designs are as follows:

Contrast Level

Line Length

B1.

B2.

B3. High

Average

Low

Medium

Treatment

A1.

5 inches

X1

x ..1

A2.

7 inches

X2

x ..2

A3.

12 inches

X3

x ..3

Jonathan Sarwono

52

Average

x.1.

x.2.

x.3.

Treatment

The design table above X1 means the respondent who gets treatment reads the

and with a 5 inch line length and a low color contrast level; X2 means the

respondent

who gets the treatment to read the ad with a 7 inch line length and medium color

contrast level and X3 means the respondent who gets the treatment to read the ad

with a 12 inch line length and high color contrast level. From the format above we

will get 9 different combinations.

CHAPTER 11

MEASUREMENT SCALE

There are four types of measurement scales in the study, namely nominal, ordinal,

interval and ratio.

Quantitative, Qualitative and Mixed Method Research Methodology

53

11.1 Nominal: The nominal measurement scale is used to classify objects,

individuals or groups; for example classifying gender, religion, occupation, and

geographical area. In identifying the above numbers are used as symbols. If we use

a nominal measurement scale, non-parametric statistics are used to analyze the

data. The results of the analysis are presented as a percentage. For example, we

classify gender variables as follows: we give the symbol number 1 and woman

number 2. We cannot do arithmetic operations with these numbers, because these

numbers only indicate the presence or absence of certain characteristics. For

example: The answer to the question in the form of two choices "yes" and "no"

that are categorical can be given numeric symbols as follows: the answer "yes" is

given the number 1 and is not given the number 2. For example in the question:

Do you agree about abortion? Answer: a. yes and b. not. If a nominal scale is used,

then "yes" is given a value of 1 and "no" is given a value of 0

11.2 Ordinal: The ordinal measurement scale provides information about the

relative number of different characteristics possessed by certain objects or

individuals. This level of measurement has nominal scale information added by

certain relative ranking facilities that provide information on whether an object has

more or less characteristics but not how many shortcomings and strengths it has.

For example: Answers to questions in the form of ratings, for example: strongly

disagree, disagree, neutral, agree and strongly agree can be given the symbols

number 1, 2,3,4 and 5. These numbers are only a symbol of rank, do not express

the amount. For example in the question: Do you agree about abortion? Answer: a.

strongly disagree, b. disagree, c. hesitant, d. agree, e. totally agree. If an ordinal

scale is used, then "strongly disagree" is given a value of 1, "disagree" is given a

value of 2, "doubt" is given a value of 3, "agrees" is given a value of 4 and "agrees

completely" is given a value of 5

11.3 Interval: Interval scale has characteristics such as those owned by nominal

and ordinal scales with added other characteristics, namely in the form of a fixed

interval. Thus researchers can see the magnitude of the differences in

characteristics between one individual or object with another. The interval

measurement scale is really a number. The numbers used can be used can be done

arithmetic operations, for example summed or multiplied. For doing analysis, this

measurement scale uses parametric statistics. For example: The answer to the

question concerns the frequency of questions, for example: How many times have

you visited Jakarta in one month? Answer: 1 time, 3 times and 5 times. Then the

numbers 1.3, and 5 are the actual numbers using intervals 2. For example in the

question: How many times did you shop at this supermarket in one month? The

answer is in the form of actual numbers: a. 1 time, b. 2 times, c. 3 times, d. 4 times

and e. 5 times

11.4 Ratio: The measurement scale ratio has all the characteristics possessed by the

nominal, ordinal scale and intervals with excess scale that have absolute 0 (zero)

empirical values. The zero absolute value occurs when the absence of a

characteristic is being measured. Ratio measurement is usually in the form of a

comparison between one individual or certain object with another. For example: A

Jonathan Sarwono

54

weight 35 Kg medium weight B 70 Kg. Then B weight compared to A weight is

equal to 1 compared to 2. For example in questions: What is your weight before

and after eating these diet drugs? The answer is the actual number: Weight before

taking medicine 70 kg and heavy after taking medicine 60 kg.

11.5 Attitude Measurement Scale

There are four scales of attitude measurement according to Daniel J Mueller

(1992), namely: 1) Likert attitude scale, 2) Thrustone scale, 3) Guttman scale, and

4) semantic differences.

a. Likert Attitude Measurement Scale: Likert scale is used to measure attitudes in

a study. What is meant by attitude according to Thurstone is "1) influence or

rejection, 2) assessment, 3) likes or dislikes, 4) positivity or negativity of a

psychological object". Usually the attitude on a Likert scale is expressed

starting from the most negative, neutral to the most positive in the following

forms: strongly disagree, disagree, not know (neutral), agree, and strongly

agree. To do quantification, the scale is then given numbers as symbols so that

calculations can be made. Generally giving the number code as follows:

"strongly disagree" given number 1, "disagree" given number 2, "do not know

(neutral)" given number 3, "agree" given number 4, and "strongly agree" given

number 5. Of course the values of these numbers are relative because they are

only symbols and not actual numbers. For Example 1: In this example a

statement that is positive, neutral and negative is written. Positive statement: I

prefer to have a Honda brand car. Neutral statement: Many Honda brand cars

are on the market. Negative statement: Honda brand cars are generally

expensive. Example 2: In this example positive, neutral and negative attitude

statements are applied in certain cases, namely in research on attitudes

towards drugs.

Commands: Choose one appropriate answer using the following answer

categories:

A. = strongly agree

B. = agree

C. = uncertain / do not know

D. = disagree

E. = strongly disagree

1. There will be no people who think healthy using drugs (N)

2. Use of Narcotics leads to the use of heroin (N) –

3. Use of Narcotics causes children born to become disabled (N) –

4. Narcotics are not "hard" drugs. (P) –

5. Narcotics have the potential for psychological therapy (P) –

Quantitative, Qualitative and Mixed Method Research Methodology

55

6. Narcotics cause a decrease in the user's humanity

Note: N is a negative attitude statement, and P is a statement of positive attitude.

The scoring process uses the following figures: or very agree worth 5, agree worth

4, uncertain value 3, disagree worth 2 and strongly disagree worth 1.

b. Thrustone Scale: The Thurstone scale is the first attitude scale developed in

attitude measurement. This scale has three attitude scaling techniques, namely:

1) pair comparison method, 2) same emergence interval method, and 3)

sequential interval method. These three methods use presumptive path

considerations that assume the relative positivity of the attitude statement

towards an object.

c. Guttman's scale: The Guttman scale is arranged based on the degree of

positivity with an emphasis on unidimensional aspects. This aspect places the

respondent at a certain point in a continuum of attitudes that must agree with

all the statement items below and must disagree with all items above the scale

position. For example: Below is an example of the Guttman scale applied in

the problem of cheating among students. The statement is composed of five

items as follows:

1) Cheating can be accepted in all circumstances

2) cheating is a habit that can be accepted among students

3) Cheating is permitted in urgent circumstances

4) cheating is acceptable if students do not study

5) cheating can be accepted if students are urged to drop out

If the respondent agrees with opinion number 1 then the person concerned

must agree with all choices under number 1. If the respondent does not agree

to statement number 1, but agrees with number 2, then he must agree with

numbers 3,4 and 5.

d. Semantic difference:

The semantic difference was invented by Osgood to measure the attributes

given by respondents to several meanings to describe certain objects. In

measuring this, adjectives are usually used which have opposite meanings.

For Example:

This example is used to measure three dimensions of meaning, namely: 1)

measuring the evaluation dimension by using four adjective pairs, 2)

measuring the potential dimensions by using as many as three adjective pairs

and 3) measuring the dimensions of activity using as many as three adjective

pairs.What is your opinion about Supermarket X?

Quick Service

---!---

!---

Slow Service

Jonathan Sarwono

56

Clean Shopping

Place

---!---

!---

Dirty Shopping Place

New Product

---!---

!---

Old Product

Low Price

---!---

!---

High Price

Spacious Parking

---!---

!---

Narrow Parking

Friendly

Employees

---!---

!---

Employee Not Friendly

Lots of Choices

---!---

!---

Little Choice

Spacious Room

---!---

!---

Narrow Room

Comfortable

---!---

!---

Not Comfortable

Safety

---!---

!---

Not Safety

11.6 Validity

A measurement scale is said to be valid if the scale is used to measure what

should be measured. For example a nominal scale that is non-parametric is used

to measure nominal variables not to measure parametric interval variables. There

are 3 (three) types of measurement validity that must be known, namely:

a. Content Validity: Content validity involves the degree to which scale items

reflect the domain of concepts being studied. A domain of certain concepts

cannot just be calculated in all its dimensions, because the domain sometimes

has many attributes or is multidimensional.

b. Validity of Construct Validity: Construct validity is related to the degree to

which the scale reflects and acts as the concept being measured. The two main

aspects of construct validity are naturally theoretical and statistical.

c. Criterion Validity: The validity of the criteria concerns the level problem

where the scale being used is able to predict a variable designed as a criterion.

11.7 Reliability

Reliability refers to the consistency and stability of the value of a particular

measurement scale. Reliability concentrates on the problem of measuring accuracy

and results

Quantitative, Qualitative and Mixed Method Research Methodology

57

CHAPTER 12

DATA COLLECTION INSTRUMENT

DEVELOPMENT

12.1 Definition

Primary data collection requires an instrument. In this section we will discuss how

to make questions using a questionnaire. In preparing the questionnaire there are

several considerations that must be done, namely: a. To what extent can a question

influence the respondent to show a positive attitude towards the things being

Jonathan Sarwono

58

asked? b. To what extent can a question influence the respondent to volunteer to

help the researcher find the things that the researcher will look for? c. To what

extent does a question dig up information that the respondent himself does not

believe in the truth?

The validity of the questionnaire is determined by the three criteria above. Besides

that the question format and answer model will also determine the quality and

accuracy of the respondent's answers. The question format is divided into two,

namely: a. How questions are asked (Question Format); b)How are these questions

answered (Answer Model).

12.2 How Are These Questions Submitted? (Question format)

The format of questions is as follows:

a. Direct Questions v.s Indirect Questions: The fundamental difference between

Direct Questions and Indirect Questions is the level of clarity of a question in

revealing specific information from respondents. Questions Directly ask specific

information directly with no direct (direct). Indirect Questions ask specific

information indirectly (indirect); however the core of the question is the same.

For example:

Direct Questions:

a. Do you like work now?

b. Do you agree with the increase in telephone rates?

Indirect Questions:

a. What do you think of the work that is currently in?

b. What do you think about the increase in telephone rates?

b. Special Questions v.s Frequently Asked Questions: Specific Questions ask

specific questions about the respondent which causes the respondent to be aware

or intrigued so that the person will give a dishonest answer. Moderate General

Questions usually ask for information sought in an indirect and general way, so

that respondents are not very aware of it.

For example: Special Questions:

a. Do you like the job of operating the production machine?

b. Do you agree with the 10% increase in Telkom's DLD telephone rates?

Frequently Asked Questions:

a. Do you like working at the company?

b. Do you agree with the increase in Telkom's DLD telephone rates?

c. Questions About Facts v.s Questions About Opinion

c. Questions About Facts will require answers from respondents in the form of

facts, while Questions About Opinions require answers that are opinions. In

practice, because the respondent may have a memory that is neither strong nor

conscious, the one who wants to create a special impression, then the Question

About the Facts does not necessarily fully produce factual answers. Likewise with

questions that ask for opinions may not necessarily produce answers that express

honest opinions. This happened because respondents distorted their opinions

based on the existence of "social pressure" to adjust to their social and

environmental desires.

Quantitative, Qualitative and Mixed Method Research Methodology

59

For Example:

Questions About Facts:

a. What brand of car do you have now?

b. What is the price of gasoline per liter today?

Questions About Opinion:

a. Why do you like Honda brand cars?

b. Why do you say the price of gasoline is cheap?

d. Questions in Sentence Form Tanya v.s. Questions in the Sentence Form

Statement: Questions in the Form of Sentences Questions provide questions

directly to respondents; while the Question in the Form of Sentences The

statement provides the answer to his agreement.

For example:

Questions in the Form of Sentences Question:

a. Do you agree with the increase in fuel prices?

b. Do you agree with divorce?

Questions in the Form of Statement Sentences:

1. The price of fuel will be raised. The answer: a. Agree b. Disagree

2. Many artists do divorce. The answer: a.Agree b. disagree

12.3 How Questions Must Be Answered

In this section we discuss the models for answering questions.

a. Unstructured Answer: This answer model is not structured usually also

referred to as open questions. This answer provides an opportunity for

residents to answer questions freely and express their opinions. The advantage

of using this answer model is that researchers can obtain complete

information from respondents; however this model has weaknesses including

the researchers will experience difficulties in processing information because

of the large amount of information data. Besides that the processing takes a

Jonathan Sarwono

60

lot of time and the researcher will have difficulty in the scoring process. For

example: Tell me how you feel about the problem of rising fuel prices What

do you think about the increase in fuel prices?

b. Field Answer: This answer model is a form of transition from unstructured to

structured question answer model. Although respondents were given

opportunity to give an open but limited response because model the question.

For example: What is your work? Which university did you graduate from?

c. Tabulation Model Answers: This answer model is similar to answer answers

but more structured because the respondent must fill in the answer in a table.

Form tables like this make it easier for researchers to organize answers

complex.

d. Answers to Scale: This answer model is another structured answer model

where respondents were asked to express their approval or acquisition to the

question given. For example 1: If you experience difficulties in what work will

you do it?

Will Stop

Working

Might Stop

Working

Take seriously but continue

to work

No

Problem

e. Answer ranking : This answer model asks respondents to rank several

statements based on their level of importance in a sequential form based on

priorities. The result is that researchers will obtain ordinal data.

For Example: ranking of these activities in relation to new product launches

Doing market research on making products, product designing, advertise

product launch products

f. Answer Form Checklist

The checklist answer asks the respondent to answer by choosing one of the

possible answers provided. The form of the answer is not in the form of scale

but in form nominal category. This form saves a lot of time for both

respondents and researchers.

Example: What type of work do you like the most?

1. Work that matches my abilities so that I can work optimally.

2. Work that forces me to work with my limited abilities.

3. work that generates a lot of money even though it is not in accordance

with my ability.