Science topic

Data Collection - Science topic

Systematic gathering of data for a particular purpose from various sources, including questionnaires, interviews, observation, existing records, and electronic devices. The process is usually preliminary to statistical analysis of the data.
Questions related to Data Collection
  • asked a question related to Data Collection
Question
6 answers
i am planning to collect blood samples from patients with a particular disease after their first visit in my department with the aim to compare the blood levels of some biomarkers with the general population. But this disease is rare, so i need around a year to collect a significant number of patients as well as data. Blood will be collected only during the first visit. As cross-sectional study by definition involves data collection at a specific point of time, what type is my observational study ?
Relevant answer
Answer
1 hypothesis about the level of antibodies
typical case-control design. The disease is known (case), and control is those who are not under disease.
2 "annual incidence of these patients for my clinic"
this is a cohort study
you observe all clinic visitors for a year and calculate the rate of visits for the alopecia
3" can I extraoplate the data for the country"
If your clinic serves a region that is homogeneous within the country - then the second question "annual incidence of these patients for my clinic" becomes "annual incidence of these patients for my region", if the region is a representative sample for the country - go ahead.
Incidence = (the number of new cases during a specific period) divided by ( total midperiod population at risk)
alopecia incidence = new cases of alopecia during one year/number of citizens in the region the same year
###
two different scientific question - two different study and design. It's OK :)
  • asked a question related to Data Collection
Question
6 answers
Brief summary of study idea
- High achieving individuals (participants)
- six month period of data collection
- 6 methods of data collection including diaries, scales, interviews and observations.
- possible sample size?
Relevant answer
Answer
There is no problem with having multiple sources of data; in fact, that can strengthen your research. The key question is what you intend to accomplish with your research. With a small sample size, you are unlikely to be able to draw statistical conclusions. This sounds like a qualitative study more than a quantitative one. Qualitative studies can provide valuable insights, which later can be tested quantitatively. Your research potentially sounds very unstructured, but that depends on how you design the interview protocols and whether you impose a structure on how the diaries are to be maintained. A risk is that you may be getting very different data from each person, with results that aren't comparable.
  • asked a question related to Data Collection
Question
4 answers
As I was reading a paper by Zattoni et al. (2013) on corporate governance theories*, I came across a passage that explains the new shift to alternative or complementary theories with the aim to explore real-life governance issues. I quote: "to enhance our understanding of the effectiveness of governance mechanisms, scholars should gain access to process-oriented data, go beyond the almost exclusive use of agency theory, and overcome empirical dogmatisms and narrow conceptualizations of corporate governance"
What does process-oriented data exactly mean? And how is it different from data used in previous corporate governance research?
* Zattoni, A., Douglas, T., & Judge, W. (2013). Developing Corporate Governance Theory through Qualitative Research: Guest Editorial. Corporate Governance: An International Review, 21(2), 119–122. https://doi.org/10.1111/corg.12016
Relevant answer
Answer
Thank you for your answer Dr. Muhammad Yousaf
  • asked a question related to Data Collection
Question
2 answers
In recent years, quite a few reports have been published of the results based on the statistical information processing. For example, a study establishes that the use of a certain remedy (some food, drink, nutritional supplement, drug, treatment method, etc.) reduces (or increases) the value of some output parameter by 20 ... 30 ... 40%. The output parameter can be the frequency of onset of the analyzed disease, the frequency of its successful cure, etc. Based on this finding the conclusion is made that the studied factor significantly influences the output parameter. How trustable can such a conclusion be?
For further details see, please,
Relevant answer
Всё зависит от конкретного исследования: иногда можно доверять, иногда нет, но даже вызывающее доверие исследование должно быть подтверждено независимыми исследованиями.
  • asked a question related to Data Collection
Question
4 answers
Can I interview a respondent who answered a questionnaires in a research at the same time?
Relevant answer
Answer
depends if the online questionnaire is following by a focus group with interview
  • asked a question related to Data Collection
Question
3 answers
Hello Good time
Please, if you have an article with the research method of grand theory and with the method of collecting information from internet websites, send it to me. Thank you
Relevant answer
Answer
Seyyed Jafar Nejat Grounded theory frequently employs the following data gathering methods: open-ended interviews with participants. Fieldwork with participants and/or focus groups Artifacts and texts are studied.
  • asked a question related to Data Collection
Question
5 answers
If the data is collected by a PG level student as an assignment for which they are to be evaluated, can this data be further used for research work by a faculty member? how do we put this in research methodology section?
Data collected by student???
Can someone please clarify this?
Relevant answer
Answer
In qualitative research the involvement of the researcher with the phenomenom has a critical roll when it comes to gathering data, therefor the vision of the student won't be the same as the one of the leading researcher. If the student is leading the research process it wouldn´t be a problem, otherwise it would be adequate to include the PG student as a fellow researcher or as an associate researcher in the project proposal.
Kind regards
Rafael Garcia Abad. PhD
  • asked a question related to Data Collection
Question
3 answers
What types of data will be required for measuring the WTP and WTA. What types of analytical model or tool be used to assess the WTP and WTA. How can I develop a questionnaire for required data collection.
Relevant answer
Answer
Thanks..@Proloy Barua
  • asked a question related to Data Collection
Question
7 answers
I am conducting a mini-research project for my master's thesis. The research study design is interpretivism philosophy. I proposed a qualitative methodological choice. I suggested using a narrative inquiry strategy in which I developed oral questions but was unsure what effective tool to use to collect the data.
Relevant answer
Answer
The following RG link is also very useful:
  • asked a question related to Data Collection
Question
10 answers
2 undergraduate researchers were tasked with doing data collection for a paper that was recently published. The data collection took months to complete. However, in the paper, there is no mention or acknowledgement of their names as the ones who collected the data.
There is a “Data Availability“ section that says that the data can be obtained by other researchers now using the same method and that the data set and analysis can be obtained via the undergraduate researchers’s professor (who is also first author). There is also an “Acknowledgements” section that says the undergraduate researchers’s professor/first author was funded by the university for completing the research (Yes, the undergraduate researchers were also granted a small amount).
Should the 2 undergraduate researchers be acknowledged in the paper for doing the data collection that was used for the paper?
Relevant answer
Answer
  • When we ask a question, we may be in doubt or don't know the answer (among others). Here I think "questioner" had been in doubt.
Reality of science dictates us that although the men/women of science generally seek to solve problems via specified methods but the path can not be the same.
All because science is decentralized, it is autonomous.
In a part of world, country, university, department one researcher, faculty member may accept a method (for example in data collection), but another in the same department does not.
So regarding the question here, my experience is this. We have undergraduate researchers that earn money for what they do (like the case of here). They have learnt how to do it fast (and may be well) and do it as a job, or part time job.
Even for many of those undergraduate researchers it is not important, they do it because they know how to do it, because they look at it as a job, in many cases for small money (as the case here again).
But the other issue is that, those undergraduate researchers are asked to help data collection by a researcher/faculty member/department/professor, or the undergraduate researchers, themselves ask (for example) their professors to help them, to learn and their names be added, for their would be future CV. Yes in cases like this, it is unethical, and even they can complain (when they have proof).
  • There is no universal answer for this question.
  • asked a question related to Data Collection
Question
3 answers
I understand that a non-sampling error is a term used in statistics that refers to an error that occurs during data collection, causing the data to differ from the true values. To what extent is it possible to quantity the extent of 'non-sampling error' to the over margin of error? Knowing fully that the overall error is function of both sampling and non-sampling error?
Relevant answer
Answer
Random measurement error (unreliability) can be quantified, for example, by using methods of classical test theory (CTT; Lord & Novick, 1968). Measurement models of CTT allow us to come up with estimates of measurement error variance components that can then be used to estimate reliability (the proportion of variance in a measured variable that is true score variance, i.e., not due to random measurement error). For CTT and related psychometric methods see, for example, Raykov and Marcoulides (2010).
There are also statistical methods for quantifying systematic method effects (i.e., non-random response "error") such as rater/observer/source effects, self-report bias, acquiescence effects, etc. in multitrait-multimethod (MTMM) research designs (Campbell & Fiske, 1959). For example, confirmatory factor analysis methods have been developed to quantify method effects in terms of variance components (so-called CFA-MTMM models; e.g., Eid et al., 2006; 2008; 2016).
Latent class (mixture distribution) analysis methods can be used to identify response styles in surveys (e.g., tendency to choose the middle category or other categories) and "filter out" individuals with a particular response style (e.g., Eid & Rauber, 2000).
Statistical methods for disentangling different variance components have also been developed within the frameworks of item response theory and generalizability theory) (e.g., https://www.tandfonline.com/doi/abs/10.1080/08957347.2011.532417.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81–105.
Eid, M. & Diener, E. (2006). Handbook of multimethod measurement in psychology. Washington, DC: American Psychological Association.
Eid, M., Geiser, C., & Koch, T. (2016). Measuring method effects: From traditional to design-oriented approaches. Current Directions in Psychological Science, 25, 275-280.
Eid, M., Nussbeck, F., Geiser, C., Cole, D., Gollwitzer, M. & Lischetzke, T. (2008). Structural equation modeling of multitrait-multimethod data: Different models for different types of methods. Psychological Methods, 13, 230-253.
Eid, M., & Rauber, M. (2000). Detecting measurement invariance in organizational surveys. European Journal of Psychological Assessment, 16(1), 20–30. https://doi.org/10.1027/1015-5759.16.1.20
Lord, F. M., & Novick, M. R. (1968). Statistical theories of mental test scores. Reading, MA: Addison-Wesley.
Raykov, T., & Marcoulides, G. A. (2010). Introduction to psychometric theory. Taylor & Francis.
  • asked a question related to Data Collection
Question
9 answers
Am conducting research on happiness at work among the teaching staff of one of the Universities in Nigeria, Enugu State University of Science and Technology in parenthesis Esut. and am looking for scale/questionnaire that can serve as my instrument for data collection.
  • asked a question related to Data Collection
Question
4 answers
I am evaluating a new English textbook. I found out the answers of teachers who are teaching this textbook in the class are very different from the answers of english teaching scholars to the same questions. How we can solve this problem at the time of giving a report to the authors of textbook to revise it?
Relevant answer
Answer
Sería muy interesante hacer una categorización de las respuestas de los docentes y del texto, de esa manera tal vez pueda observarse que tan disímiles son y ambas partes trabajar en el acercamiento.
  • asked a question related to Data Collection
Question
2 answers
I am looking for some programme that allows me to identify bot accounts in the data collected by me.
Relevant answer
Answer
Identifying bots is one of the most prevalent industry practices, and many tools are available for this. One such tool is FollowersAnalysis.
There are various tools that can help you in the identification of fake followers. One such excellent tool is FollowersAnalysis. They specialize in studying Twitter followers and their usage patterns which allows them to separate the needles from the haystack.
Getting a detailed account analysis is the most effective solution available today to maintain the authenticity of a Twitter account.
Although Twitter has taken steps to curb this issue but considering the size and reach of Twitter, it is easier said than done.
  • asked a question related to Data Collection
Question
8 answers
Dear researchers,
Recently i have conducted a online survey work through google form.I want to know what could be it's methodology?
Help me suggesting any good articles where authors used/mentioned this types of survey process.
Thanks in advance!
Relevant answer
Answer
I agree with the previous thoughtful responses. I want to stress that as an alternative questionnaire instrument you may prefer the "Qualtrics.com" over Google form (It requires a subscription though). I found it is easy to get a better response since this platform provides a wider possibility of presenting your survey questions. Non-response rate might affect this kind of survey very much.
About the methodology, in addition to other ideal procedures in any other research, please consider some points:
-Develop your research question wisely based on a review of existing literature in line with the question of your interest
-If there is a control group, the characteristics should be commonly shared between the treatment and control groups.
-Close-ended questions are very helpful to reduce data management efforts after completing the data collection part.
-The selection of the study population needs to be validated in a systematic way for example who you choose to participate in the research. Questionnaires should be randomly distributed over any network.
I hope it helps.
  • asked a question related to Data Collection
Question
7 answers
Hello everyone,
This might be a very common question/challenge about quasi-experimental research. I have been doing a data collection for my study. In the end, the data is not adequate (i.e., the participants' participations were declining over the course of five weeks. Only 10 out of 39 participants had the interventions for five weeks).
Some said that I can still run a statistical analysis with some adjustment on p-value and specific analysis (but I think this might not be something that journal editor would expect). Some other said that I could pivot the study to a more qualitative study with additional work of interview.
Any suggestions based on your experience? That would be much appreciated. Thanks
Relevant answer
Answer
if you cant manage to find another target group to collect data from please use secondary data from similar research
  • asked a question related to Data Collection
Question
1 answer
Can anyone please explain how to analyse data collected using brief cope 28 item scale?
Like i won't get the total score rather than score for each subscale and then it is mentioned in a research to use normative data of a heart failure study for calculating percentile ranks. Can anyone please help as I dont get the data analysis part after collection using this scale?
  • asked a question related to Data Collection
Question
5 answers
Dear Research Gate community,
I’m conducting a study on how different management practices in wetland (ponds) affects the diversity and abundance of species in the wetlands. The different ponds are next to each other. Some ponds are control without any management practices. The water in the treatment ponds is regularly drawn down and filled up again with water from the river. We recorded the waterbird species number and abundance of each pond regularly (record the bird data of all the ponds at the same time for each survey).
- In the first year, we conducted a baseline study in which no treatment was done for all the ponds (data were collected monthly).
- In the second year, we conducted the treatment (operational study), and data of birds were collected weekly.
We’re now trying to study:
1) first, if there is any difference between the treatment ponds and control ponds during the operation
2) if there is any difference between baseline study and operational study of the same pond.
We wonder what kind of statistics are suitable for statistically analysing our data.
Some problems we are encountering is:
1. The data do look like normally distributed. The data collected are time series data, there are natural seasonal variation in the number of waterbirds in our region (a lot of migratory birds in fall and winter). How to take into account of influence of the time of survey.
2. The sample frequency for baseline year (12 times) and operational year is different (52 times), how to compare the difference between baseline and operational year.
Highly appreciate any help or suggestion!
Best regards
Relevant answer
Answer
Try to check a book on: Design and Analysis of Experiments, just be sure that you are doing the right design. Regards.
  • asked a question related to Data Collection
Question
3 answers
Can anyone recommend a usable resource/tool for species detection from acoustic data please? A middle ground between a phone app and something like Arbimon would be about the level. Briefly, a phone app lacks flexibility in data collection, i.e. it cant be left out all night or left running for long periods. However, Arbimon is not that useful to anyone below ecologist level, as it only tells the user what species is present if the user completes their own validation, i.e. the user has to identify all species themselves. I'm looking for something that can analyse data from an AudioMoth, uploaded by citizen scientist participants, and actually identify species.
Relevant answer
Answer
Very interesting project.
There also are some other interesting findings when doing a google search for "training data animal sounds"
  • asked a question related to Data Collection
Question
3 answers
How do I justify scientifically a small sample size for data collection for quantitative data collection to validate my findings for the low response?
Relevant answer
Answer
Hello Christtestimony,
I agree with Abdelrahman M. Farouk that it appears you've posed virtually the same query recently. So, at least some of the answers there are applicable here.
In a nutshell, when you are faced with a low response rate, the chief concern is whether your sample is in any dependable way representative of your intended target population. (A secondary concern is that of statistical power or the precision of any parameter estimates you might wish to make.)
What you can do, if information is available, is compare your obtained sample to the target population on a number of obvious characteristics. For example: How does the age distribution of your sample compare with that of the target population? Distribution by sex? By education level? and so on... You can conduct goodness-of-fit tests to see whether your sample varies to a statistically discernible degree from the target population. If it does not, then you have better hope that the variables of interest to your study are perhaps plausible estimators than if your sample is known to differ from the population on such characteristics.
Note: This does not guarantee representativeness, but it's better than nothing.
As well, you can always go out and collect additional data, but that is frequently perceived as a less-than-desirable solution!
Good luck with your work.
  • asked a question related to Data Collection
Question
20 answers
It may help the researcher to drop or add the necessary questions before going to the main data collection work. It is a preliminary feasibility research process that will help the researcher to make his research more soulful. It will help to adjust the research question and research design and to prevent the unnecessary cost and time expenditures in the main data collection.
Relevant answer
Answer
Thank you Professors and Development Professionals for having answered my question.
Pilot study will help the researchers or scholars to have an exact questions pertaining to the title but addition and deletion of the questions possible only after conducting the pilot study only.
  • asked a question related to Data Collection
Question
6 answers
Dear researchers,
I am a coordinator of an international research project investigating the impact of remote and hybrid work on employees' wellbeing, retention, and knowledge transfer. Currently, we want to proceed with data collection in Taiwan, Hong Kong, and Singapore (we have prepared the survey in Chinese traditional). However, this is rather complicated for my team (even partners from HK and Taiwan) to find relevant data collection sources (websites/companies/Universities) that could help in getting the right respondents. I posted a batch on Amazon MTurk, but gather only a few relevant answers.... Could you share any ideas or contacts with companies/websites helping in data collection in mentioned countries?
Relevant answer
Answer
Please can you give me the link of Kantar company.
  • asked a question related to Data Collection
Question
8 answers
I plan to collect data for my research through two questionnaires. The first questionnaire is for collecting data regarding the two independent variables from managers, and the second questionnaire is for collecting data about the dependent variable from customers (Customers' responses). So, the question now is how to handle this in the statistical analysis, given that the number of questionnaires collected from customers will be greater than that will be collected from managers (about 350 managers and 500 customers)? Also, is it ok if we have inequivalent numbers of questionnaires from managers and customers, or is it a must to insert the same number of responses for each variable in the statistical program (such as SPSS) to statistically examine the relationship between them?
Thank you so much in advance.
Best regards.
Relevant answer
Answer
I believe that the number of collected surveys depends on what variables and what industry are you interested in. However, you can find formulas about the minimum sample size on the web.
this study could give you a good insight into the survey structure and the minimum sample size formula.
  • asked a question related to Data Collection
Question
2 answers
I am conducting a study to understand the factors that will impact usage mobile health application. Your participation in this study is completely voluntary. All the information obtained will be kept strictly confidential and anonymous. The data obtained will be only for research purposes. Please select the most suitable response as applicable to you. https://forms.gle/x7kTK9mh6s1FPXgB7
Relevant answer
Answer
Thank you so much
  • asked a question related to Data Collection
Question
4 answers
Dear Scholars & Researchers,
How do I justify scientifically a small sample size for data collection for quantitative data collection to validate my findings for the low response?
Looking forward to hearing from you all.
Relevant answer
Answer
Dear Christtestimony Jesumoroti
Regrading the sample size, it would kindly suggest to follow the following:
1) There are many software that calculates the minimum sample size
2) It depends on your study filed and objectives
3) The sample size can be small and still valid as scholars generally agree that a sample size of 30 is sufficient for statistical data analysis and the drawing of meaningful conclusions [1]
4) Relate with past or similar studies done on your filed area, this can increases the reliability of your work, there are few studies that had small sample size as in [2,3]
5) The sample size should reach saturation which can indicates that the sample size is enough
6) The objectives of the study should be addressed through the given sample size and if the sample size is small it can mentioned as a limitation [4]
References
[1] Ott, R.L.; Longnecker, M.T. An Introduction to Statistical Methods and Data Analysis; Cengage Learning: Boston, MA, USA, 2015
[2] Osei-Kyei, R.; Chan, A.P.C. Developing a Project Success Index for Public–Private Partnership Projects in Developing Countries.J. Infrastruct. Syst. 2017, 23, 04017028. [CrossRef] [3] Adabre, M.A.; Chan, A.P.; Darko, A.; Osei-Kyei, R.; Abidoye, R.; Adjei-Kumi, T. Critical barriers to sustainability attainment inaffordable housing: International construction professionals’ perspective. J. Clean. Prod. 2020, 253, 119995. [CrossRef]
[4] Tah, J. H. M., Thorpe, A., & McCaffer, R. (1994). A survey of indirect cost estimating in practice. Construction management and economics, 12(1), 31-36.
  • asked a question related to Data Collection
Question
2 answers
Hello,
To be sincere, I do not know if I can write this here, but I will try.
Some time ago, I have been working in order to understand and identify the drivers behind Land Concentration in Chile. Even though I have made progress, I am looking for a quantitative expert as a coauthor today.
The principal issue will elaborate an econometric model using the data collected.
If someone is interested just let me know.
All the best
Relevant answer
Answer
Hello
What type of data do you have?
Crossectional or longitudinal?
  • asked a question related to Data Collection
Question
8 answers
Dear all,
For my graduation research, I am trying to create a composite score about household resilience out of data collected through a household survey. However, this data consists of ordinal variables (5-point Likert scale), binary variables (yes-no questions), and ratio variables (in proportions between 0-1).
My plan was to recode the data on the 5-point Likert scale into scores from 0-1 and do the same for the yes-no questions, with yes = 1 and no = 0 (since answering yes would mean a household is more resilient and thus have a higher score). However, this seems very off.
At this moment I am aware that it wasn't the best idea to create a survey with both types of questions, but I am unable to recollect the data.
Therefore my **question** is as follows: do you have any tips on how to create a composite score composed of both interval and binary variables?
Thank you in advance and have a lovely day.
Best,
Nina
Relevant answer
Answer
Hello Nina,
In general, I would not recommend collapsing a measure having more than two values down to just two values unless there was some very compelling reason to do so. You're jettisoning potentially useful information by so doing.
The question here is, how do you intend to use the composite score you seek to construct? That should help to discern appropriate approaches from inappropriate methods. The fact that some elements of that composite are Likert-type response scale values whereas others are dichotomous variables isn't a deal-breaker.
When you say that you have "interval" variables, are you using individual Likert-type items as variables, or have you created and vetted one or more unidimensional scale scores by summing/averaging/otherwise combining scores over a set of related items? Generally, most would agree that individual Likert-type items involving self-report of human perception/sentiment/opinion/judgment (et al.) are ordinal strength scores, not interval strength scores. There are ways to get the results across related sets of such items into interval strength scores, however.
Good luck with your work.
  • asked a question related to Data Collection
Question
11 answers
I have data from a question asked to people pre- and post-treatment and answered on a 10 point Likert scale. I need to analyse the change in response. Searching for an answer has proven much more difficult than I expected, as there seem to be a lot of different tests used in subtly different circumstances. Can anybody give me some guidance on what the best test to use is in this situation?
Relevant answer
Answer
The paired samples Wilcoxon test (also known as Wilcoxon signed-rank test) is a non-parametric alternative to the paired t-test used to compare paired data. You could use this for each of your two variables.
  • asked a question related to Data Collection
Question
6 answers
I have a model with the following parameters:
  • Groups: factor - 4 levels (base level = control group)
  • Time: numerical
  • Label = factor - 3 levels (base level = control group)
  • price = numerical (5 different values; from a Likert scale)
The problem is that the base level of variable 'Groups' is perfectly collinear with the variable Time because, in the control condition, no values for Time were collected. That means that for Groups = 'control group', Time is always '0'.
This introduces singularities or perfect collinearity in my regression model, meaning I cannot interpret it correctly.
Do you have any suggestions for helping me out? Recollecting data is not an option, unfortunately. It was too costly and there is too many time constraints.
Relevant answer
Answer
Time isn't collinear because it hasn't been measured. Time is not known for the control group, not zero. So you cannot use the time variable. You can examine the other variables, but since you have no data for the controls you can only examine the effect of time within the intervention groups.
  • asked a question related to Data Collection
Question
10 answers
Actually, I have one independent variable and a dependent variable, I am using one latent variable which is a mediator and through that I wanted to know the effect on dependent variable. The problem is I am not aware if data collected for latent variable is through semi structured questionnaire and not on likert scale will make the possibility for Structure Equation Model? Kindly explain and help. I am rookie researcher and new to multivariate analysis and other methods.
Relevant answer
Answer
This would require the open ended questions to reflect something that can be meaningfully scored on a continuous (quantitative, i.e., at least ordinal) scale. If the answers to those open questions are purely qualitative in nature (nominal scale), it may be difficult or impossible to include those variables in an SEM as indicators of continuous latent variables.
  • asked a question related to Data Collection
Question
7 answers
this my data collection form any one you have opinion about added or remove some things and thank you for all
Relevant answer
Answer
Life style
  • asked a question related to Data Collection
Question
4 answers
The costs and time commitments associated with data collection and labeling might be prohibitive. A huge dataset is insufficient since the success of deep learning models is strongly dependent on the quality of training data. Cost, time, and the use of appropriate training data are all challenges. Biases, incorrect labels, and omitted values are some of the difficulties that impair the quality of deep learning training datasets.
Relevant answer
Answer
I agree, but they're other metrics that can be deployed to appraise these algorithms apart from accuracy.
  • asked a question related to Data Collection
Question
3 answers
One dependent variable (continuous) ~ two continuous and two categorical (nominal) independent variables
I'm seeking for the best method for predicting a data collection with more than 100 sites. The distribution of all continuous variables is not normally distributed.
Relevant answer
Answer
Beyond the scarcity of information, are you sure of the relationship between variables?
  • asked a question related to Data Collection
Question
2 answers
I received this from our ethics committee "Declare potential COI of the principal investigator to the research participants," I'm a bit confused of what I should put. My research is all about Coping Strategies of Primigravida Teenage Childbearing Mothers in Metro Manila, and the purpose is to determine the coping strategies of the respondents and correlate it with their demographic profile.
Relevant answer
Agree with Carl L Tishler that every journal has its own instruction on where to put the COI. It's usually in the instruction for authors. When it's unclear, I usually put it in the cover page, which is separated from the body of the article. This is for the purpose of anonymous review process since the COI may reveal your indentity. Afterwards when the review is completed, you may put the COI in the body of text (e.g. after conclusion, before reference list) or as the journal directed. But where you put the COI is not a deal breaker if you put it at the wrong place. The journal's administrator will inform you to move it elsewhere before passing it to the editor.
  • asked a question related to Data Collection
Question
11 answers
regarding my project, I am wondering if it would possible to you to help me find the method to calculate sample size and data collection strategy in a qualitative study? also, questions to include and a timeline for evaluation.
Thanks
Relevant answer
Answer
Qualitative research emphasizes and develops a detailed understanding of participants’ experiences. It is unconcerned with sample representativeness; therefore, it does not apply a priori sample size estimation—as in quantitative inquiry. Instead, it employs non-probability sampling methods (e.g., convenience sampling, quota sampling, & snowball sampling). In other words, individuals are chosen for the sample because they can furnish elaborate inputs to address the questions under investigation. A common principle for determining the adequacy of qualitative data is saturation, which refers to the notion that additional data collection or analysis is unnecessary based on the data that has been already gathered or analyzed. The following can render germane guidance.
Saunders, B., Sim, J., Kingstone, T., Baker, S., Waterfield, J., Bartlam, B., Burroughs, H., & Jinks, C. (2018). Saturation in qualitative research: Exploring its conceptualization and operationalization. Qual Quant,52(4), 1893–1907. https://doi.org/10.1007/s11135-017-0574-8
Sim, J., Saunders, B., Waterfield, J., & Kingstone, T. (2018). Can sample size in qualitative research be determined a priori? International Journal of Social Research Methodology, 21(5), 619–634. https://doi.org/10.1080/13645579.2018.1454643
Good luck,
  • asked a question related to Data Collection
Question
11 answers
Data is collected from 2 Sources:
1.Source A
2 Source B
For Source A, Data on the topic~subject is collected over a very long period of time on an individual.For Source B,Data on the topic~subject is collected over a relatively brief period on Millions of people.
1.What is or shall be the Principles for Quantitative Determinations of the Reliabilty of Data collected from Source A and Source B?
2.How does one assess and evaluate the Authenticity of the Principles in 1 ?
Relevant answer
Answer
Reliability is a property of the data you collect, rather than the source of the data -- although the nature of the source can affect how reliable the data is.
The reliability of data of data often thought of in either of two ways. The first is test-retest reliability, so that measuring the same variable twice produce similar results each time. The second is the degree of random error, because rand om error can correlate with anything, so the more random error a measurement contains, the less reliable it is.
  • asked a question related to Data Collection
Question
4 answers
Dear Colleagues
I need to find the data (dengue cases in Thailand for every province,2003-2009).
Does anyone know if this information is publicly available?
Thank you very much
Relevant answer
Answer
Let me share my experience. I collected secondary data from Proncial Health Office (PHO) in Thailand for my PhD research. They require ethical approval. if you live in Thailand, my suggestion is to go to a nearest PHO and check whether dengue cases data are available for each province. You need a local guide if you cannot speak Thai.
  • asked a question related to Data Collection
Question
3 answers
Hi. Everyone. I have a problem with WordSmith tools on doing the data collection of 4-word Lexical Bundles that I only see the number over there but cannot check the data source like I use AntConc to do 4-word bundles that can be shown the LBs types. So how can I check the data source of 4-letter words? For example, Text 1 has 135 LBs was calculated and where can I see the whole LBs types? I spent a whole day finding out, but nothing can help at all. I am so confused about this tool now. If anyone can help me, I will appreciate it.
Relevant answer
Answer
Laoxin Qin To begin creating a word list, push the WordList button on the main Controller. When WordList loads, choose your texts, and you'll see something like this. Here, we'll create a single simple wordlist based on eight text files from the play Romeo and Juliet, so click Make a word list immediately.
  • asked a question related to Data Collection
Question
1 answer
Self-reporting bias is a challenge in case of GPS data collection, in studies where participants have to manually START and STOP recording their trip, unlike studies where GPS data is passively collected (continuously in the background) without the need for user intervention.
I specifically want studies which have mentioned the existence of self-reported bias in the context of GPS data collection.
Relevant answer
Answer
I'm still trying to understand your question. In the passive collection, you are referring to a GPS device that is at a fixed location so the deviations in reporting can be seen?
And you are looking for what?
  • asked a question related to Data Collection
Question
5 answers
Hi
in my SEM model, i managed to have the attached value with low loadings but good model fit. Could any body provide any insights on if this results are valid for publication or still need improvement.
PS: I really struggeled in DATA collection & the model is only for testing can't be edited.
Sample size :66
CMIN: 1,052
GFI: 0,882
AGFI:0,82
CFI: 0,939
TLI:0,92
RMSEA:0,28
Thanks
Relevant answer
Anass Ziat , Great question to follow.
The question is exciting, but it may be too late to answer. I agree with other professors that the sample size (n=66) is too small. The minimal required sample size for inferential statistics should be over 100. It is perfect if it is over 400.
Also, some factor loadings are lesser than 0.7. Experts should prove the measurements for content validity before data collection.
All the best,
  • asked a question related to Data Collection
Question
5 answers
Dear scholars, I have a technical question that needs clarification. As employees and customers are the most participants when it comes to the data collection from the questionnaires related to organizations or companies studies. Therefore, there may be some reasons or justifications for the choice to include both employees and customers in the same sample without employing multi-groups analysis. Besides, I am not sure that employees and customers points of views are homogenous and that both can be considered as part of the same population. So, my question is what are the reasons that justify that the employees and customers sample size (i.e. data collected from the questionnaires) can be combined to form the same population of the study? In the other words, from which extend, employees and customers can be homogenous to form the same population of the study?
Relevant answer
A good question to follow. You have to ensure the same questionnaire is suitable for employees and customers. Pre-testing could benefit in this case. However, the demographic characteristic could separate employees and customers. The advantage of this study, you may study the difference in perceptions among employees and customers, it is depending on your research objectives. Thus, the homogenous is feasible.
All the best,
  • asked a question related to Data Collection
Question
4 answers
Hi everyone,
I have a three-group experiment design with two data collection times (baseline and retention). Some data violate the assumptions of normality and homogeneity of variance. But for the data that meets assumptions, I am using repeated measures ANOVA to show three points:
First, all subjects are coming from the same population (using the normality test of Shapiro-Wilk [also skewness value] and homogeneity of variance of Levene's test). Second, the impact of treatment (between-subjects), and third, if there is a significant difference between baseline and retention sessions where the significance lies (within-subjects).
Besides normality and homogeneity of variance tests, I am not conducting any other tests (like unpaired t-tests) for the first point. I am conducting ANOVA for the second and third points. I believe this is correct but I would like to confirm this?
My main question here is; what are non-parametric tests that I need to conduct to show the same three points for the parametric tests?
Here is what I am planning and would appreciate any feedback or any suggestion on what to use.
I am planning to run a set of pairwise Mann-Whitney U tests to show that all three groups are coming from the same population. Use Friedman's test to show the impact of treatment. Then run another set of pairwise Mann-Whitney U tests to show where the differences are. Is this the right approach?
Thank you and appreciate your help!
- Hawkar
Relevant answer
Answer
You're pretty confused. See a statistical consultant at your university. Best wishes David Booth
  • asked a question related to Data Collection
Question
3 answers
Should I discard the absorbance less than 0.1 and above 1 when making a calibration curve?
Relevant answer
Answer
I don't think you should discard them. These limits help you to determine your limits of detection and quantification. The typical response is a linear curve; but if you plot your curve and realise that the response is not linear, you can use other functions. I always use linear curves though. I do sometimes have to remove portions of the curve that are non-linear, then include a linear regression equation which I use to estimate the quantity of analyte in my sample.
As long as the concentrations of the standards are in a range that includes the concentration of the analyte of interest, you are good.
  • asked a question related to Data Collection
Question
14 answers
How to maintain data quality in qualitative research? How to ensure quality in qualitative data collection as well as data analysis?
Relevant answer
My pleasure, professor Dr Devaraj Acharya .
Kind Regards,
  • asked a question related to Data Collection
Question
3 answers
Dear scientific community,
Based on reviewers suggests, i am considering inter-rater analysis, therefore measuring 2nd time, for my study. However, there are some aspects are not clear;
1) For the second measurements; should i measure the variables of whole subjects or a part of them would be sufficient (i have seen in some articles 20% of the variables were remeasured and analyzed) ?
2) İf whole data should be remeasured instead of a part of them, should the results of 2nd. measurement included to 1st measurements' results by averaging the results of two measurements which also means i have to rewrite results and discussion section?
3) or remeasuring a part of the subjects variables and presenting the results of 1st measurements which already given in the article would be correct?
Relevant answer
Answer
Can you say more about things were measured?
  • asked a question related to Data Collection
Question
2 answers
Currently my study is focussed on 'Understanding Customer Behaviour under Interest Rate Changes'.
I would like to know the list of types of financial data that are aggregated nationally. Total deposits is one such example.
Where can I find the data of 'Total deposits' and the 'other financial data' that are aggregated nationally.
Please suggest the official websites for gathering the above data (not kaggle, github etc).
Relevant answer
Answer
you will contact statistical data web site
  • asked a question related to Data Collection
Question
6 answers
We have survey 15 teachers and their students (6 - 14 students each) to find a correlation between teacher well-being and students' academic results. We are also looking at student-reported teacher relationship as a potential mediator.
Unfortunately, we have had to stop data collection due to COVID19.
We are considering using Spearman's think that it is most likely not sufficient. Is there a better alternative for this?
Should we also drop mediation analysis given the sample size obtained?
Relevant answer
Answer
Hong Hwee Note that your data have a multilevel/hierarchical structure (students nested within teachers). This is relevant for statistical inference (significance testing) and needs to be taken into account, for example, by using multilevel/hierarchical linear modeling to analyze the data. That being said, you have relatively few Level-2 units (teachers) which may limit your power.
  • asked a question related to Data Collection
Question
2 answers
Hello,
I am in need of ECG and EMG data collected for the same person for to create ML models?
Could you provide pointers to that data collected for set of people.
Relevant answer
Answer
Dear Kumar ,
There are many ECG and EMG repository are available online openly. some of the link listed
Also you can directly contact and request to the corresponding author of journal related to your work.
  • asked a question related to Data Collection
Question
4 answers
We plan to distribute supply chain surveys via personal networks. Do you have any experience to share? I think the point(s) of penetration is important. Credible people and credible websites support the data collection. I plan to use LinkedIn and Facebook.
Relevant answer
Answer
Dear Robert,
I posted a link to the survey form created in the Google form on my Facebook and LinkedIn profiles. For now, I have few completed surveys, because I only recently posted this link. I am planning a post with a link to the survey form also on my profiles of other social media portals, e.g. on Twitter. In my opinion, it is also worth considering posting a post with a link to the survey form also on other social media portals, on various discussion forums and on scientific portals, e.g. on the Research Gate portal. The question of which social media portals, websites, discussion forums, scientific portals, etc. should contain links to the survey form, depends on the scope and specificity of the research subject to which the survey relates, and also on whom we address the questions in the survey survey questionnaire.
Best regards,
Dariusz
  • asked a question related to Data Collection
Question
7 answers
There are different guidelines for what qualifies a researcher to be an author on publications. I've just seen a case on Twitter where the PhD student was puzzled that the supervisor was going to publish an article from his/her data collection without including him/her.
For PhD students: do you think this is fair? have you discussed it with your supervisor?
For PhD supervisors: What criteria do you use for establishing the authorship policy in your lab/research group?
I see this a huge gray zone in research, and clarity/transparency would help everybody :)
Thanks!
Relevant answer
Answer
Sadly, you are right, Robert: authorship rules in a research group is a huge "gray zone" which lies outside the legal field. Usually, one person decides who and with whom publish. In fact, the Good Scientific Practice are just recommendations.This means: whatever happens, it can be not nice at all but is not against law... Another important issue is, it is impossible to prove that someone's name was "forgotten" on purpose. Of course, as long as the society sticks to the presumption of innocence. The classical excuse is like: "my student edited the manuscript and forgot your name (or was innocently unaware), and, by the way, he does not work in my group anymore ...
  • asked a question related to Data Collection
Question
9 answers
The data collection in my study will be analysed using the Activity theory which is a qualitative data analysis method. Just want to find out which research approach is suitable for the study? Is it inductive or deductive approac?
Relevant answer
Answer
mixed approach will provide good output
  • asked a question related to Data Collection
Question
7 answers
Hello, I'm an urban design graduate student and it's my first semester. How is research done? I need to write an example at the process of learning to identify a Problem, Problem, Method. Research Proposal: Tension in green urbanism (or strategies): Displacement communities in large-scale projects (İzmir - GreenUp?)
Research Problem: Displacement of community in the implementation of green strategies in cities / Socio-spatial flaws in green strategies against climate change
Research Questions: Are necessary preventive measures taken against displacement of communities for large-scale green strategies? Can the scope of green strategies be both environmentally beneficial and socially equitable?
What I am wondering is that the Greenup project for Turkey is considered new and not completed. We are expected to learn and do until the data collection and method determination part in order for it to be empirical research. I wonder how can I do it, your suggestions and comments are important to me, thank you!
Relevant answer
Answer
There are many different activities and topics associated with green strategies & displacements communities (DC). Such as CC & DC include sea level arising inundation and DC, farmland degeneration and DC, flood & DC etc; hydropower and other renewable energy (as CC mitigation approach) and DC; ecological conservation and DC; environmental protection and DC......
you can find many articles in my academic home (Guoqing Shi in researchgate, email: gshi@hhu.edu.cn) regarding climate migration, dam/hydropower induced displacement and resettlement, environmental migration and so on.
We have Master/Ph.D programs in Resettlement Science and Engineering, in (Resettlement ) Sociology, in Public Administration since 1993/2004 and in English since 2013 in Hohai University Nanjing China (www.hhu.edu.cn). You are welcome to study here in future.
Myself, I engaged in Ilisu dam as international resettlement and social expert during 2006-2009.
you can pay attention to
0. Doi: 10.1126/science.1208821
1. DOI:10.1080/14615517.2021.1980277
2. DOI:10.1080/14615517.2021.1980277
3. Doi: 10.1126/science.1208821
4.doi:10.1080/14615517.2019.1706386.
5.doi:10.1007/978-3-642-23571-9_10
6.doi:10.3390/su11072142
7.doi:10.1007/978-3-030-57426-0_19e
8.doi:10.1080/14615517.2020.1798722
9.doi:10.1177/1070496511426478.
10.doi:10.1080/07900627.2016.1216829.
11.doi:10.1080/07900627.2017.1417824.
12.Doi:10.4324/978/1003159780
13.doi:10.1007/s10661-017-5967-6.
14.DOI: 10.1111/jfr3.12609
15.doi:10.3390/su14052488
16.doi:10.3390/su11072142.
  • asked a question related to Data Collection
Question
9 answers
  • What is the difference between Landsat Collection 1, Collection 2 Level 1and Collection 2 Level 2?
  • What are the specifications included in each type?
  • Do these types include atmospheric correction?
  • What is the most appropriate collection for Land Use landcover and ecological related analysis?
Relevant answer
Answer
Dear Janaki,
Maybe you would be able to find your answer in the following USGS link:
  • asked a question related to Data Collection
Question
1 answer
I would like to know, if single investigator (specialist) will assess different variables on x Ray.
Does he requires to do inter examiner reliability test with expertise before I start data collections , or he needs to do only the intra examiner reliability test for all variables? Because all variable will be assess by single investigator? On the other hand, the work will be done by single author.
Relevant answer
Answer
If all of the data are assessed by a single investigator, then you are limited to intra-rater reliability. In general, this is not recommended because if that rater has some kind of systematic bias, then that will simply be repeated in an intra-rater scoring.
  • asked a question related to Data Collection
Question
5 answers
I have species abundance (different sp.) data collected from 6 sites at two different seasons. I want to test if there is seasonal variation in the bundance of species by using the Wilcoxon signed-rank test (paired data). Can anyone help me?
Relevant answer
Answer
Manova?
  • asked a question related to Data Collection
Question
2 answers
If before, then 2 samples from the same population are needed, one to establisg reliability of the instrument and another one for the actual study. Need some clarity about the right approach.
If it has to be done before the actual data collection then how many participants should be an ideal number for just establishing the instrument's reliability..
For a validated instrument (designed in consultation with the subject matter expert), what if it is done after the actual data collection and the reliability coefficient (CronBach Alpha) falls below the acceptable range.. what should be done in such scenario..
Relevant answer
Answer
Thanks David for respondng,
Actually , I am not using an existing instrument, It's a new instrument designed in consultation with a subject matter expert for a new study with different sample..
If I expose the instrument to the sample in pre test and then if the reliability coefficient falls below acceptable range then I will have to drop some questions/items to establish the reliability.. e.g if instrument has 10 items, i might have to drop 2 or 3 or 4...
That's one approach..
Or I run the the instrument through a separate sample from the same population and establish the reliability before the actual intervention .....
I do not see any clear directions in this regard...
  • asked a question related to Data Collection
Question
5 answers
Hello everyone,
Does anyone have any idea, is there any website or people who help in data collection for research article professionally?
If yes, I would request to share the contact details ?
Relevant answer
Answer
Kindly specify the kind of data that you want to collect
  • asked a question related to Data Collection
Question
4 answers
I am currently working on my dissertation and would like input on the best platform to use for a study that includes data collection at two points in time and is not cost prohibitive. What platform would you recommend?
Relevant answer
Answer
Hello Debbie,
An electronically distributed/collected approach is by far the most economic method for data collection (once you're past any cost outlays for the software, of course) .
The two concerns you'll need to address are:
1. If your aim is to match scores by respondent across the two occasions, you'll need some sort of identification information. This often makes getting approval from IRB committees tougher, as you're departing from anonymized measures.
2. If your query is really about, which online survey platform should I use, then have a look at this comparison summary: https://www.pcmag.com/picks/the-best-online-survey-tools
Good luck with your work.
  • asked a question related to Data Collection
Question
5 answers
We intend to write a review paper with data from various studies around the world, but the data collection in the various studies was done using different methods and intensities. So, in this case, how should the data set be standardized in order to run some analysis?
Relevant answer
Answer
I'm not sure if this can be done even when the raw data for all included studies is available to you; the impact of sampling type/effort in the collected data may be hard to disentangle even with the use of statistical assumptions that are unlikely to have been validated previously.
If you really think that sampling type and sampling effort are really important factors to consider in the phenomenon you are interested in, you may consider separating the pooled analyses per these factors. You may also opt to not do quantitative analyses and focus on qualitative synthesis.
  • asked a question related to Data Collection
Question
5 answers
I'm currently looking at the rheological properties of the polymer Xanthan Gum. focusing on its dynamic viscosity to be more specific. I'm assessing the effects of pH (ranging from 3.6 to 5.6, 0.4 increment, total of 6 pH's) on the dynamic viscosity of xanthan gum solution (dissolving xanthan gum powder into acetic buffer with equal ionic strength, concentration is kept at 0.04%).
Firstly, my viscosity data collected shows that, as pH increases from 3.6 to 4.0 then 4.4, the viscosity increases; but as I bring up the pH from 4.4 to 4.8, 4.8 to 5.2, then lastly 5.2 to 5.6, the increasing viscosity trend plateaus and the increase in viscosity is less significant compared to the 3.6-4.4 jump. At this range, does pH has an effect on the viscosity of xanthan gum based on its molecular configuration? Though some sources states that xanthan gum's viscosity remains stable and unchanged within the range of pH 3-12 at a high concentration like 1% not 0.04%, yet some suggest pH still plays an effect, though I'm not sure how on the chemical and molecular aspect.
A possible conjecture I can think of is the xanthan gum's order-disorder and helix-coil transition is affected by protonation. In figure 2, it demonstrates how electrolytes affect the structure of the polymer; in figure 3, it shows how at a state of a helical rod and no longer a random coil, it is capable to hydrogen bonds among each other. Hence, I'm wondering of pH plays an effect on it's structural transition, such that the increased intermolecular forces at the form of a helical rod would make it more viscous in solution.
Here are the resources I have used so far:
Brunchi, CE., Bercea, M., Morariu, S. et al. Some properties of xanthan gum in aqueous solutions: effect of temperature and pH. J Polym Res 23, 123 (2016). https://doi.org/10.1007/s10965-016-1015-4
Relevant answer
Answer
Dear Ryan Lo, you may find various stufies on this topic. Please have a look at the following free access RG fille. My Regards
  • asked a question related to Data Collection
Question
3 answers
Recently, an acquaintance of mine told me about a friend of hers who earns money by responding to digital surveys sent to her by a data-collection company such as Prolific or Amazon Mechanical Turk.
Apparently, when answering surveys the woman's friend does not concentrate on the responses she provides (often watching TV simultaneously) but primarily ensures she answers a sufficient number of questions in a way that appears valid so that she can claim payment for having submitted her responses.
I therefore wonder about the extent to which data collected that way is valid. Can anyone provide insights, or experiences, concerning this, please?
Relevant answer
Answer
If possible, scrutinise some random datasets.
  • asked a question related to Data Collection
Question
4 answers
Many young researchers are faced with uncertainty when they go from the step of developing and planning their methodology to acutally doing the data collection. Even professors often cannot predict well, how long data collection takes.
The data collection phase therefore can substantially impact the total duration of a research project.
I believe it should become a standard to record the time needed for field work and lab work and publish these figures along with the actual research.
This would enable other researchers to plan their campaigns better and it would have the side effect of making the effort and time going into research much more visible.
It would also make planning recurring data collection and inventories easier to plan and budget, which would in the long run contribute to more realistic expectations of the actual work volume and effort that researchers will be faced with.
It can also help prevent going on unrealistic or overambitious data collection campaigns and make it easier to select appropriate measurement and field methods.
In terms of recurring inventories, it would result in opportunities for contractors to better estimate working and setup times for making realistic offers and simultaneously helps contracting bodies to give contracts to those bidders who plan budget realistically.
What do you think?
How could such a standard be implemented?
Relevant answer
Answer
You are right that the "data collection phase ... can substantially impact the total duration of a research project" and that even "professors often cannot predict well, how long data collection takes". However, I doubt that collecting data on the time the data collection phases took will help to improve the situation. Each investigation in each field of science has its own problems - travelling to remote areas for geological, biological or archaeological field work, doing experiments in high-energy physics or observations in astronomy, collecting interviews in psychology, reading historic manuscripts, etc. Even in one and the same field each study should include new ideas and cannot be compared completely to previous studies.
  • asked a question related to Data Collection
Question
3 answers
I am embarking on a research project and want help with a data collection instrument on the emotional and academic performance of visually impaired students who are the target population.
Relevant answer
Answer
Hi Bro,
The following article could possibly give you an insight for measuring emotions
  • asked a question related to Data Collection
Question
16 answers
Some literature clarify that they used quarterly data from the source of world-bank is that available? or they transform the annual data? how this transformation done?
Relevant answer
Answer
International Debt Statistics • These include the high frequency, quarterly data for high-income economies and select developing countries reporting to the joint World Bank–IMF Quarterly External Debt Statistics (QEDS) and the Quarterly Public Sector Debt (PSDS) database. [on-line] https://data.worldbank.org/products
• Quarterly Public Sector Debt (QPSD) [on-line] https://www.worldbank.org/en/programs/debt-statistics/qpsd
  • asked a question related to Data Collection
Question
3 answers
Data collected using Likert Scale and collecting insights from Experts, entrepreneurs and customers
Relevant answer
Answer
In order to do CFA, you have to a prior model that assigns each item to a specific conceptual factor. If you do not have a prior model, then you can use EFA to develop one.
  • asked a question related to Data Collection
Question
3 answers
I am working in the field of data discovery, therefore, I learn data collection using API of social apps.
Relevant answer
Answer
Lee Mccallum Thank you!!!!!
Sales Aribe Jr. Thank you!!!!!
  • asked a question related to Data Collection
Question
3 answers
Mixed method research design.
Relevant answer
Answer
The Triangulation Design (Figure 4.1a) is the most frequent and well-known technique for mixing procedures (Creswell, Plano Clark, et al., 2003). This design's goal is to "collect distinct but complementary data on the same issue" (Morse, 1991, p. 122) in order to better comprehend the research challenge.
It entails gathering both types of data at roughly the same time; assessing information using parallel constructs for both types of data; analyzing both types of data separately; and comparing results using procedures such as a side-by-side comparison in a discussion, transforming the qualitative data set into quantitative data.
Exploratory sequential mixed methods, on the other hand, is a way of integrating qualitative and quantitative data gathering and analysis in a series of phases (Creswell & Plano Clark, 2018).
  • asked a question related to Data Collection
Question
5 answers
The popularity and applications of the metaverse are expected to skyrocket in the near future, thanks to the entry of major players (Facebook, Microsoft, EA, and even McDonalds) in the industry. As ab opinion piece in The Drum puts it "many experts predict that the metaverse will change the way we live and work in the future, and it would be extremely negligent not to recognize the trend and continue to engage with it". Could someone suggest some preliminary studies that could be carried out using consumer data collected through online surveys? Thanks in advance.
Relevant answer
Answer
Chances are that technology adoption and acceptance theories will be soon applied to figure out what might be the antecedents for individuals to use the metaverse. I just hope that the operationalization of the respective constructs will be done in a meaningful way; otherwise we will just get the usual and fairly tautological results saying that "people will use it because it is useful". Having said that, it might be interesting to investigate privacy concerns and potential privacy violations. This can be done with or without technology acceptance models.
  • asked a question related to Data Collection
Question
5 answers
I have collected data at 3 different time periods. For the 1st time period, I have analysed using SEM (SPSS + AMOS). Subsequently, I have collected data for 3 different time periods (say each year). What type of statistical analysis can be done for drawing inference ?
Relevant answer
Answer
Why you needed to collect data at three different time periods? Your answer lies in there!
  • asked a question related to Data Collection
Question
5 answers
Hi, I am a PhD student working on my thesis about game platforms and human behavior. I am having a hard time at the data collection stage and would like to ask for some suggestion and advise. I started distributing my questionnaire on various online forums (e.g. Reddit) and social media (Facebook) dedicated in gaming but with low response rate. Is there any other forums or website that I can distribute my questionnaire online and reach more participants for data collection?
Relevant answer
  • asked a question related to Data Collection
Question
5 answers
Hello everyone. I am a PhD student at the Czech University of Life Sciences, I am working on the development of data collection collars (tags) for different animals species (and, what is more critical - sizes). At this point, I am looking for colleagues, coauthors and collaborators who are working in the field of animal behaviour and automatic data collection (i.e. accelerometers, gyroscopes, GPS). To continue the development and appliance of animal behaviour data collection devices and analysing tools for that purpose. If you are interested in more details, you can contact me here or by email: fiialkovskyi@ftz.czu.cz
Relevant answer
Answer
Dear Pavlo,
Thanks for your prompt reply, so let us start up our joint work on animal behaviors. I am looking forward to hearing from you on this affair very soon.
Thanks in advance,
Takeo
  • asked a question related to Data Collection
Question
23 answers
I am working on a project related to online abuse. The structured questionnaire consists of about 50 questions and we would like to pretest the survey questionnaires before actual data collection. What would be the appropriate way to determine the sample number for pretesting?
Relevant answer
Answer
Since the population is 50 take 10 percent of it for the pretest.
  • asked a question related to Data Collection
Question
5 answers
Hello researchers,
Sometimes, we collect a huge amount of raw data during field visit, that we are unable to use in a research article immediately. But there are possibilities to use that data in future. So, I would like to know for how many years primary data collected from the field by the individual researcher will be valid. For example, can I use the data I had collected in 2015 through field survey, to write a paper in 2022? Will it be still valid?
Relevant answer
Answer
Dear Payel,
I hope it depends on the data type but mostly the primary data's of more than 5 year old are not recommended.
  • asked a question related to Data Collection
Question
5 answers
Dear Sir/Mam,
For analysis can we take companies listed in nse, and two or three indices? Or should we just take listed companies from only one indices. Kindly tell. Thank you.
Relevant answer
Answer
Refer to these articles:
Firm Value and Cross-Listings: The Impact of Stock Market Prestige
Analytical Study of Correlation between Indian and International Stock Market