Is it a must for a quantitative study to have hypotheses?

Hi Prabhashini, 

I think the sense of formulating hypotheses is to test existing theories and make predictions about these are, in very general and absolute terms, true or fake. The subsequent quantitative analysis is only one way by which you can do above. A quantative analysis without explicit hypotheses is only a descriptive analysis, probably without any aim....you simply describe a reality but you cannot make any inference or comment any theory... 

You may consider to have some readings on epistemological issues.

Regards, Fabio.

Hi Prabhashini, 

I second what Fabio La Rosa states, and would add that any statement of p-value or probability implies that some null-hypothesis is being tested. In other words, if you are using any statistical tests, then hypothesis testing is implied.  Presumably, it should be stated, but often it is not.  

Hi Prabhashini, 

If you are interested in purely qualitative/text data, you can apply Grounded Theory. Please try with NVIVO software to gather knowledge on how to implement  this.

THanks all of you for the answers. They are of immense value

As Fabio notes, you could limit yourself to purely descriptive work, but then the issue would be whether that makes a substantial contribution to your research field.

Hi Prabhashini,

Hypothesis represents the important step in the scientific method of research, so it is important to formulate your hypothesis according to your question in research problem and the test of the hypothesis may  give the answer to your question.

Good Luck

Hi,

No, it is not a must to have hypotheses in all quantitative research. Descriptive studies dont need hypotheses.

however, RCT and experimental studies, require having hypothesies,

and when you want to use inferential statistics also you need.

Regards

Thanks a lot for your answers

HI

Prabhashini Wijewantha

I think you have sufficient answers to your question. I will further add that sometime when we do not have sufficient previous knowledge in form of literature to base our hypothesis on, then we can also use a research question instead of hypothesis and then use inferential statistics.

Regards.

Thanks a lot Muhammad Imran for your answer. I am so sorry that I am replying you very late. I actually went through your answer,but I could not respond at that time as I was in a hurry. Thanks once again.

Dear Prabha.....It’s not compulsory in all cases that you must have a hypothesis. But for the clarification of your proposed outcomes it’s better to develop one hypothesis, only in the case when you will use inferential statistics for achieving your objectives.

It is possible to do quantitative research without hypotheses. For a research objective such as "to explore and describe the intensity of Green HRM practices being applied by state corporations in Sri Lanka" you don't basically need to formulate a hypothesis and you can achieve the objective by using descriptive statistics (simply mean, standard deviation, range etc). However, if you want to formulate a hypothesis in this regard, it is possible, and that type of hypothesis is called Descriptive Hypothesis. For example, "Intensity of Green HRM practices being followed by state corporations in Sri Lanka is at good level." It is possible to use One Sample Technique to test the hypothesis.

If you use NGT (Nominal Group Technique) to identify and prioritize reasons, or aspects for/of a certain managerial phenomenon, you can use weights and percentage analysis (without using a hypothesis). 

Best wishes to you, Prabhashini.

Dear Sir,

Thanks a lot for the detailed reply which is clearly understandable because of the examples that you have used. Thanks once again

 "Hypothesis is used in quantitative research (The Research Process, 2016 p. 38).".  

Thanks everyone for the valuable insights shared in the form of replies

I think it is usual to have hypothesis with quantitative research, but what is more important is that before deciding on whether your study needs a hypothesis or not, you need to answer the following questions:

–Does your study take an experimental approach to answer questions?

–Are you making a prediction about the phenomenon being studied?

Goodday all.

Mine is not to suggest an answer to the existing issue, but to ask a similar question on the research I am undertaking now. I want to know whether or not a research tilted::

needs hypothese?

Dear Hussaini

If this is a pure descriptive study providing tables of information on needs, resources and service provisions and you do NOT plan to do any association studies then I do not need any need for any null hypothesis. You could still have research questions. You can state this as a descriptive study and hypothesis generating one. However if you plan to see if the resources or information provided has an impact on the function of Khadis etc then you need hypotheses and apply inferential statistical methods to analyse your data etc.

I agree, no need of an hypothesis if the research question calls for a description of a situation or of some aspect of reality

on which the research is focused.

Once a research have research question there is no need for any hypothesis. However, exploratory or descriptive research may not necessary need hypothesis. The question for quantitative or qualitative research has no relationship with hypothesis formulation.

Thank you for all yours answers.

I also have the same question on the research I am undertaking now. My research is "Develop theoritical basis for the creation of web atlas (collection of maps) of National Parks"

Does it need hypothese?

Definitely no need for a hypothesis in a descriptive study. Descriptive studies are sometimes used to explore a given phenomena with the purpose of developing a hypothesis. Once we describe a phenomenon in a descriptive study (e.g. we find that a high percentage of people who eat bread seem to have cancer), we can test any developing hypothesis in an experimental study later (e.g. "Bread causes cancer" might be one hypothesis). So in short, descriptive/observational studies don't need a hypothesis. Experimental designs (e.g. RCT, CCT, Repeated measures) need a hypothesis to test.

Phuong Van - It doesn't sound like you're testing a hypothesis. Is it a quantitative study?

Coefficient alpha is used to assess reliability. It is based on the average correlation among a set of items, with high correlations indicating that the items consistently measure the same underlying concept.

I am not sure what you mean by "descriptive" questions, but if the goal is to determine whether there is agreement across the questions (i.e., the answers are highly correlated), then alpha would be appropriate.

This will depend on what type of study you're doing. If it's descriptive only, you're not testing a hypothesis. In this case, there's no reason to use cronbach's alpha.

can we write assumptions instead???

can we write assumptions instread???

There can be proposed and emerging hypothesis for a descriptive or experimental research design. When the research question needs to verify the existing finding taken from a similar study done with different unit of population and wants to verify with another population using a quantitative study. We propose a hypothesis and verify using inferential statistics. Emerging hypothesis are derived as the researcher during the analysis of data finds a trend of relationship between variables and contributes as his new finding for the future researchers in the similar research in future.

I would recommend you this paper: Malhotra, M.K. and Grover, V., 1998. An assessment of survey research in POM: from constructs to theory. Journal of operations management, 16(4), pp.407-425.

Las investigaciones cuantitativas exploratorias y descriptivas puras no necesitan hipótesis. Cuando las investigaciones descriptivas predicen o pronostican una cifra o un hecho, solamente en estos casos necesitan una hipótesis.

Claro que se puede trabajar sin hipotesis, pero exclusivamente segun define Hernadez Sampier, en investigaciones cuantitativas descriptivas,prospectivas o retrospectivas; en el momento de describir, tenemos al objeto de estudio frente a nosotros, lo observamos y las caracteristicas son evidentes, pero en el caso de una investigacion cuantitativa exploratoria, o en un correlacional ya no vemos todas las variables u objetos de estudio, ahi si necsitamos una hipotesis, un pronostico, una@ proyeccion de un posible resultado.

You may check more details about this query at:

> https://conjointly.com/kb/hypotheses-explained/

> https://www.researchgate.net/post/Without_hypothesis_can_researcher_conduct_survey_for_phd_thesis

Cuando se realizan investigaciones cuantitativas exploratorias o descriptivas puras, no requieren plantear hipótesis. En el primer caso (exploratorias) vamos a investigar algo nuevo del cual no conocemos a profundidad o conocemos poco sobre la variable, y no vamos a demostrar nada. En el caso de las descriptivas puras no necesitan hipótesis porque se van a describir fenómenos ya existentes; a menos que vayamos a predecir o pronosticar un hecho. Solo en este caso planteamos una hipótesis.

Yes, you need a hypothesis if you are doing a quantitative study because you don’t know the answers. If you know the answer why do a quantitative study? Any empirical research needs a hypothesis to test the results.

El momento que planteas preguntas para encontrar respuestas estas planteando una investigación correlacional, entonces debes tener una hipótesis, pero si vas a describir un fenómeno existente pero desconocido en la población investigada, no necesitas hipótesis porque no vas a demostrar nada, simplemente a describir dicho fenómeno.

It depends on the nature of the research questions. However, inferential statistics demand hypothesis testing.

Being able to formulate hypotheses indicates that a certain level of experience has been acquired, then with the rapid development of the times and our increasing ignorance of the future, fewer and fewer people are able to initiate studies with huge investments through self-sufficient authoritative hypotheses, which creates a major obstacle to research.

Also, hypothesis may be considered as a direction of effort to be established in quantitative research, but why wouldn't it be an inherent prison of thinking? So there is a need for some expansion and change for the research method itself. And I think what matters is not whether there is a hypothesis, but whether the expected outcome, no matter what (no advance hypothesis), is very important to the knowledge we need to discover.

If you need to answer a research question, then you need to state your null hypothesis.

Quantitative research - Is there minimum number of Hypothesis required?

Hola. El enfoque de una investigación, sea cuantitativo, cualitativo o mixto, no determina la existencia o no de una hipótesis. El alcance que tenga la investigación es el que orienta al planteamiento de una hipótesis. Las investigaciones con alcance correlacional y explicativo, siempre llevarán hipótesis. Las de alcance exploratorio no llevan hipótesis, las de alcance descriptivo no llevan hipótesis, a menos que intenten predecir o pronosticar un dato o un hecho.

If you have research questions, then it might be essential to also have a hypothesis. However, not all quantitative studies strictly require hypotheses. Some exploratory or descriptive research might not involve explicit hypotheses but instead focus on analyzing patterns and relationships. Additionally, in certain types of research, such as correlational or cause-and-effect studies, the emphasis might be on examining links between variables without necessarily making specific predictions.