Why does data collection stall so many dissertations?
Collection is where a dissertation timeline most often loses months, even when every earlier chapter is finished on schedule.
- Ethics approval takes longer than planned. A form returned for revision can add weeks that were not built into the original timeline.
- Recruitment is slower than the literature makes it look. Wu et al. (2022) put the average online survey response rate in education research at 44.1%, so on average more than half of the people invited do not answer.
- A sampling plan written after the fact does not hold up. An answer decided after the data came in reads as an afterthought when a supervisor asks why the sample was drawn that way.
- Raw responses need cleaning before they mean anything. Incomplete entries, duplicates and inconsistent coding have to be resolved first.
What is MAAS Data Collection Support?
The stage between an approved ethics application and a dataset ready for analysis: instrument design, sampling plan, real responses or a secondary dataset, then cleaning. No target grade, quoted by scope; MAAS EdTech does not fabricate data or fill in responses to reach a target sample size.
Who is Data Collection Support for?
Students and researchers whose question and ethics approval are settled and who now need the instrument built and the data gathered, the design steps Crotty (1998) places at the methods layer.
- Dissertation and thesis students with an approved ethics application who now need the questionnaire or interview guide built
- Researchers who need a sampling plan they can defend when asked why the sample was drawn that way
- Students who need a secondary dataset sourced because primary collection is not feasible
- Anyone with raw survey or interview data that needs cleaning before analysis
This is not for you if your ethics approval has not yet been granted. If your ethics form itself is not yet written, an expert can guide you through preparing it, which adds 5% to the base price, quoted separately.
What do you get?
An instrument, a sampling plan, a dataset built from real responses, and the documentation that shows how it was gathered.
Instrument design
- A questionnaire or interview guide built from your research question, checked against the target population it measures, so what counts as a good response is clear before recruitment starts
- A pilot round where useful, to catch confusing items before full recruitment. Van Teijlingen and Hundley (2001) list 16 reasons to pilot, from testing whether an instrument is adequate to checking that the sampling frame and the recruitment approach work
Sampling and recruitment
- A sampling plan written before recruitment starts, with the reasoning documented for your methodology chapter, the layer Crotty (1998) and Staller (2013) place sampling decisions at
- Recruitment follows the protocol your ethics board approved. Where that protocol names only you as the researcher, the expert prepares the invitation, instrument and sampling plan, and you send them out and collect responses yourself
- A secondary dataset sourced and matched to your research question, when primary collection is not the right route
Data cleaning
- Incomplete and duplicate entries flagged and resolved by a documented rule, not removed at random
- Coding checked for consistency against a documented rule. Van den Broeck et al. (2005) recommend that a study report its cleaning method, the error types and rates found, and how many errors were deleted or corrected, so each rule is written down for your methodology chapter
- A file handed back in the format your analysis stage needs
How strict are ethics timelines in practice?
Ethics rules are set programme by programme, so check your handbook for when the ethics form is due and how it counts towards the methodology chapter before any primary or secondary data is touched.
Data protection law sits alongside ethics review. For research in the UK, Article 89 of the UK GDPR makes the research provisions conditional on appropriate safeguards, and the Information Commissioner's Office (n.d.) advises using anonymous data where the research allows it and pseudonymising it as early as possible where it does not, because pseudonymous data is still personal data.
Whichever route your data takes, primary collection, a case study applying an existing framework, or secondary dataset reanalysis, it has to match the epistemological stance argued earlier in the methodology chapter (Crotty, 1998).
Where does an instrument and a sampling plan actually sit in your methodology?
Crotty (1998) separates the research process into four nested layers: epistemology, theoretical perspective, methodology, and methods, placing statistical analysis, interviews, questionnaires and case studies at the methods layer, the layer a sampling plan and a data collection instrument belong to. Staller (2013), summarising Crotty (1998), situates survey research, ethnography, grounded theory and action research one layer up, at methodology. Not every author agrees on where each technique sits: Creswell and Creswell (2023) group quantitative, qualitative and mixed methods at the methodology layer instead of splitting them across layers, and McCarthy (2012) notes that Crotty treats a case study as a method while other authors treat it as a research strategy. Saunders, Lewis and Thornhill (2012) warn that many texts use "methodology" and "methods" interchangeably regardless of which convention they claim to follow, so keeping a sampling plan clearly at the methods layer is easier for a supervisor to follow.
Why does a documented sampling plan hold up better under questioning?
A sampling plan written down before recruitment starts holds up better under a supervisor's questions than one reconstructed afterward, because it states the reasoning before anyone has to ask for it.
How many participants is enough has published evidence behind it rather than a rule of thumb. For interviews, Guest et al. (2006) coded 60 in-depth interviews and found that 92% of the codes they built across the first 30 transcripts had already appeared within the first 12 interviews, with the basic elements of their metathemes present as early as 6. Hennink and Kaiser (2022) reviewed 23 studies that tested saturation and found it was usually reached within 9 to 17 interviews or 4 to 8 focus group discussions, mainly in studies with homogenous populations and narrowly defined objectives, while the few outliers, such as multi-country research, needed larger samples. For a survey that estimates a proportion in a defined population, Krejcie and Morgan (1970) published a table built on 95% confidence and a 5% margin of error: their worked example of 9,000 high school teachers needs 368 responses, and the figure levels off at 384 for a population of 1,000,000. For a survey that tests an effect, the target comes from a power calculation: Cohen (1992) set the small, medium and large effect sizes (d = 0.2, 0.5 and 0.8) that researchers fall back on when no better estimate exists, and Faul et al. (2007) describe G*Power 3, the free program many students use to run the calculation. The response rate then decides how many people to invite. Wu et al. (2022) screened 8,672 studies and examined 1,071 online survey response rates in education research: a clearly defined population raised the rate, while sending the survey to more people did not and incentives showed no significant effect. Hickey et al. (2018) add that real studies lose participants to withdrawal, loss to follow-up and missing data, so the number recruited has to exceed the number the calculation asks for.
What happens to the data after analysis, and who collects it?
Nothing further, unless you ask for it. Once your dataset moves into SPSS, R, or Python analysis, you keep the raw file, the syntax, and every output, so you can explain how each result was reached. Your researcher works in your discipline, because a sampling frame that fits a marketing survey does not fit a clinical interview study. Level, field and country are visible on the experts page, and how engagements have run on success stories.
References
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159. https://doi.org/10.1037/0033-2909.112.1.155
Creswell, J. W., & Creswell, J. D. (2023). Research design: Qualitative, quantitative, and mixed methods approaches (6th ed.). SAGE.
Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research process. SAGE.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146
Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59–82. https://doi.org/10.1177/1525822X05279903
Hennink, M., & Kaiser, B. N. (2022). Sample sizes for saturation in qualitative research: A systematic review of empirical tests. Social Science & Medicine, 292, 114523. https://doi.org/10.1016/j.socscimed.2021.114523
Hickey, G. L., Grant, S. W., Dunning, J., & Siepe, M. (2018). Statistical primer: Sample size and power calculations, why, when and how? European Journal of Cardio-Thoracic Surgery, 54(1), 4–9. https://doi.org/10.1093/ejcts/ezy169
Information Commissioner's Office. (n.d.). What are the appropriate safeguards? https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/the-research-provisions/what-are-the-appropriate-safeguards/
Krejcie, R. V., & Morgan, D. W. (1970). Determining sample size for research activities. Educational and Psychological Measurement, 30(3), 607–610. https://doi.org/10.1177/001316447003000308
McCarthy, J. (2012). Reflections on business research (RIBM Working Paper Series). Manchester Metropolitan University. https://mmu-eprints-repo-prod.mmu-eprints.cdl.cosector.com/593399/1/JeffMcCarthy-ReflectionsOnBusiness-Research-RIBM-Mar2012.pdf
Saunders, M., Lewis, P., & Thornhill, A. (2012). Research methods for business students (6th ed.). Pearson.
Staller, K. M. (2013). Epistemological boot camp: The politics of science and what every qualitative researcher needs to know to survive. Qualitative Social Work, 12(4), 395–413. https://doi.org/10.1177/1473325012450483
Van den Broeck, J., Cunningham, S. A., Eeckels, R., & Herbst, K. (2005). Data cleaning: Detecting, diagnosing, and editing data abnormalities. PLoS Medicine, 2(10), e267. https://doi.org/10.1371/journal.pmed.0020267
van Teijlingen, E. R., & Hundley, V. (2001). The importance of pilot studies. Social Research Update, 35. University of Surrey. https://sru.soc.surrey.ac.uk/SRU35.html
Wu, M.-J., Zhao, K., & Fils-Aime, F. (2022). Response rates of online surveys in published research: A meta-analysis. Computers in Human Behavior Reports, 7, 100206. https://doi.org/10.1016/j.chbr.2022.100206
