Chapter 4
Data Analysis
Masters Dissertation
PhD Thesis
Kenyan Universities
SPSS
Research Methods
Dissertation Defence
Common Mistakes Students Make in Chapter Four Data Analysis — and How to Fix Them Before Your Supervisor Does
Tobit Research Consulting | Postgraduate Research Skills Series | Reading time: ~18 minutes
What you will learn: Why Chapter 4 is the chapter that most often stalls a dissertation or thesis at the correction stage at Kenyan universities; the specific data analysis mistakes that supervisors and examiners flag most consistently across KU, UoN, JKUAT, MKU, Strathmore, Moi, Egerton, and Laikipia; how to present descriptive statistics, inferential statistics, and qualitative findings in the way your panel actually expects; how to interpret and discuss your results rather than merely reporting them; how to maintain the alignment between Chapter 4 and the objectives you committed to in Chapter 1; and the pre-submission checklist to work through before you hand Chapter 4 to your supervisor.
Students who reach Chapter 4 have already done something genuinely difficult: they have conceptualised a research problem, reviewed a substantial body of literature, designed a methodology, collected data in the field, and survived the administrative gauntlet of Kenyan university postgraduate processes. By the time they sit down to write the data analysis chapter, many feel the hardest work is behind them. It is not.
Chapter 4 — Data Analysis and Presentation of Findings — is the chapter where the scholarly promise of your earlier chapters is either fulfilled or exposed. It is where examiners find out whether you actually understand the methods you cited in Chapter 3, whether you can distinguish between reporting results and interpreting them, whether your analysis truly answers the research questions you committed to, and whether you have the analytical rigour that postgraduate-level research demands.
The feedback patterns that supervisors and examiners consistently produce on Chapter 4 submissions across Kenyan universities are not random. The same mistakes appear repeatedly — in dissertations from different institutions, different disciplines, and different research approaches. They are predictable, which means they are avoidable. This guide addresses each of them directly.
At Tobit Research Consulting, we work daily with Masters and PhD students across Kenya — at Kenyatta University, the University of Nairobi, JKUAT, Mount Kenya University, Strathmore, Laikipia University, Moi, Egerton, and Kisii University — reviewing Chapter 4 drafts and helping students transform raw output tables and interview transcripts into polished, examiner-ready findings chapters. What follows is a direct account of the mistakes we see most often, and precisely what to do about each one.
1. What Chapter 4 Is Actually Required to Do
Before examining the specific mistakes, it is worth being precise about what Chapter 4 is and is not. Chapter 4 is not a storage location for your SPSS tables. It is not a transcript of your interviews. It is not a summary of what your questionnaire showed. It is an analytical narrative — a chapter that presents your data in organised, reader-accessible form; interprets what that data means in relation to each research question; and begins the scholarly conversation between your findings and the theoretical and empirical literature you reviewed in Chapter 2.
At most Kenyan universities, Chapter 4 carries a standard structure: an introduction, a response rate section, a pilot study results summary (where applicable), a demographic profile of respondents, objective-by-objective presentation and interpretation of findings (for quantitative studies), thematic analysis of qualitative data (for qualitative and mixed-methods studies), and a chapter summary. The chapter ends where Chapter 5 — Discussion, Conclusions, and Recommendations — begins. Understanding that boundary is itself a source of many Chapter 4 failures.
The central standard your examiners apply to Chapter 4: Every finding must be presented clearly, interpreted accurately, and traced back to a specific research objective. If a table or result appears in your chapter that does not correspond to a stated objective, it is noise. If an objective is stated in Chapter 1 but generates no finding in Chapter 4, it is a gap your examiner will not overlook.
2. Mistake 1: Inadequate or Missing Response Rate Analysis
Mistake 1
Common Error
Skipping the response rate, reporting it without comment, or misunderstanding what it means
Many students either omit the response rate section entirely, mention it in a single sentence (“The researcher distributed 200 questionnaires and received 178 back”), or report a percentage without explaining whether it meets the threshold for valid analysis. Some students calculate the response rate incorrectly, confusing distributed questionnaires with the target sample.
The response rate section is not a formality. It is the first thing an examiner reads to assess whether your data collection was sound enough to draw the conclusions you are about to draw. At Kenyan universities, a response rate of 70% or above is generally considered acceptable for quantitative research, with 80% and above regarded as strong. Rates below 60% require explicit justification — including a discussion of potential non-response bias and how it might affect your findings.
✔ How to Fix It
Report the response rate as a percentage with the numerator and denominator clearly stated and displayed in a table. State whether the rate meets the threshold recommended in your methodological literature (cite the source — Mugenda & Mugenda, 2003, is widely used in Kenyan university research for this purpose). Briefly discuss what the response rate means for the representativeness of your data. If your rate is lower than ideal, acknowledge it and explain the likely causes — tight data collection windows, organisational access restrictions, low digital literacy among respondents — and address what, if any, effect this may have on the generalisability of your findings.
✍️ Weak vs. Strong Response Rate Reporting
Weak: “Out of 200 questionnaires that were issued, 178 were returned, giving a response rate of 89%.”
This reports a number. It does not say whether 89% is adequate, why it matters, or what it tells the reader about the data’s reliability.
Strong: “Of the 200 questionnaires administered to the target sample, 178 were returned fully completed, yielding a response rate of 89%. This exceeds the 70% threshold recommended by Mugenda and Mugenda (2003) as the minimum acceptable rate for meaningful statistical analysis, and is considered sufficient to draw valid inferences about the study population. The 11% non-response rate is attributed primarily to respondents who were unavailable during the data collection window due to operational commitments at their respective institutions.”
3. Mistake 2: Reporting Descriptive Statistics Without Saying Anything About Them
Mistake 2
Common Error
Presenting frequencies, means, and standard deviations in tables — then moving to the next table without interpreting what the numbers show
The most persistent Chapter 4 error across Kenyan university dissertations is the presentation of descriptive statistics tables followed by prose that simply restates what the table already says: “Table 4.3 shows that 45% of respondents agreed, 30% strongly agreed, 15% disagreed, and 10% were neutral.” This is not analysis. It is transcription.
Descriptive statistics — frequencies, percentages, means, standard deviations, and modes — describe the shape and central tendencies of your data. Their purpose in Chapter 4 is to give the reader a clear picture of what your respondents indicated and to set up the inferential analysis that follows. But the description of what the numbers show is the student’s job, not the table’s.
✔ How to Fix It
For every descriptive statistics table, write a paragraph that does three things: identifies the key finding (what does the majority pattern or mean value indicate about respondents’ experience or opinion?); flags notable variations or extremes within the data (were there any items where opinion was highly divided, or where agreement was near-universal?); and connects the descriptive result to the objective it addresses. A mean score of 3.8 on a Likert scale of 1–5 tells you something — but only if you say what a score of 3.8 in your measurement scale means. Does it indicate agreement? Moderate satisfaction? Relatively high perceived effectiveness? Interpret it.
✍️ Descriptive Interpretation — Weak vs. Strong
Weak: “Table 4.5 shows that the mean score for staff training was 3.92 with a standard deviation of 0.74.”
Strong: “As shown in Table 4.5, respondents rated staff training as a component of credit risk management with a mean score of 3.92 (SD = 0.74) on a five-point Likert scale, indicating general agreement that training programmes are in place and perceived to be effective. The relatively low standard deviation suggests reasonable consensus among respondents on this item, with limited dispersion around the mean. This finding indicates that staff capacity development is an established practice within the study institutions, though the mean falling below 4.0 (the ‘strongly agree’ threshold) suggests room for improvement in training quality or frequency.”
4. Mistake 3: Using the Wrong Statistical Test for Your Data Type
Mistake 3
Common Error
Running Pearson correlation on ordinal Likert data, applying regression to nominal variables without encoding, or using t-tests when ANOVA is required
The choice of statistical test must match the measurement scale of your data and the nature of the relationship you are examining. This is not a technicality — it is the methodological foundation on which your findings rest. An examiner who identifies a test applied to the wrong data type will question not just that finding but the credibility of your entire analysis chapter.
🇰🇪 Test Selection at Kenyan University Level — What Examiners Expect
The table below summarises the most common data types encountered in Kenyan postgraduate research and the appropriate tests for each. Your Chapter 3 committed to specific tests — Chapter 4 must use exactly those tests, or explain (with methodological justification) why a different test was substituted.
| Data Type / Research Question |
Appropriate Test(s) |
Common Mistake |
| Relationship between two continuous variables |
Pearson correlation (if normally distributed); Spearman rank correlation (if non-normal or ordinal) |
Using Pearson on raw Likert items without testing normality |
| Difference between two group means |
Independent samples t-test (two groups); paired t-test (same group at two time points) |
Using t-test for three or more groups (ANOVA required) |
| Difference between three or more group means |
One-way ANOVA; post-hoc Tukey or Bonferroni if significant |
Running multiple t-tests instead of ANOVA (inflates Type I error) |
| Influence of multiple independent variables on one dependent variable |
Multiple linear regression (continuous outcome); logistic regression (binary outcome) |
Using correlation where regression is needed; omitting regression assumptions tests |
| Association between two categorical variables |
Chi-square test of independence |
Using chi-square when expected cell frequencies fall below 5 (use Fisher’s exact instead) |
| Comparing an observed distribution to expected |
Chi-square goodness of fit |
Applying a two-variable chi-square test to a one-variable distribution |
✔ How to Fix It
Before running any inferential test, confirm: What is the measurement scale of each variable involved (nominal, ordinal, interval, ratio)? Is the data normally distributed — and have you tested this using Shapiro-Wilk or Kolmogorov-Smirnov? How many groups are being compared? Is the research question about association, difference, or prediction? Every test you run in SPSS, Stata, or R must be justified in Chapter 4 with a sentence that explains why that test is appropriate for the data type and research question at hand.
5. Mistake 4: SPSS Output Dumping — Copying Tables Without Interpretation
Mistake 4
Common Error
Copying SPSS output tables directly into the dissertation without editing, labelling, or interpreting them
Raw SPSS output is not dissertation-quality presentation. It is a working tool for the analyst. When it appears unedited in a dissertation — with default SPSS formatting, unexplained column headers like “Sig. (2-tailed)” or “Std. Error Mean,” and no surrounding interpretation — it signals to the examiner that the student used SPSS without understanding what the software produced.
✔ How to Fix It
Every table in Chapter 4 must be reformatted to meet your institution’s table standards (APA 7th format is most commonly required: no vertical lines, clear title above the table, notes below). Extract only the relevant statistics from the SPSS output — do not paste the entire correlation matrix if only two variables are relevant to the objective you are discussing. Label every column and row in plain English. Below or alongside the table, write an interpretive paragraph that explains what the statistical result means: the direction of the relationship (positive or negative), its magnitude (weak, moderate, strong), its statistical significance (p-value relative to your alpha level of 0.05), and its meaning in the context of your research question.
✍️ Interpreting a Regression Result — What Your Examiner Expects
Weak (output dump): “The regression results are shown in Table 4.7.”
[Table appears with no interpretation. Student moves to next section.]
Strong: “Table 4.7 presents the results of the multiple linear regression analysis conducted to examine the influence of credit risk assessment practices on loan default rates among youth borrowers. The model was statistically significant (F(3, 174) = 18.42, p < .001) and explained 24.1% of the variance in loan default rates (R² = .241), indicating that the independent variables together account for a meaningful, though not exhaustive, proportion of the variation in the outcome. Among the predictors, collateral assessment (β = −.38, p < .001) emerged as the strongest significant predictor of reduced default rates, followed by borrower financial literacy screening (β = −.27, p = .003). Loan tenure flexibility was not a statistically significant predictor (β = .09, p = .214), suggesting that repayment scheduling alone does not independently reduce default risk within this study population.”
6. Mistake 5: Confusing Findings Presentation With Discussion
Mistake 5
Common Error
Either keeping Chapter 4 entirely bare of interpretation (leaving all discussion for Chapter 5), or burying Chapter 5-level discussion inside Chapter 4
This is one of the most structurally confusing aspects of the Kenyan university dissertation format, and it produces two opposite errors. Some students present raw results in Chapter 4 with no interpretive voice at all, intending to “discuss” everything in Chapter 5 — leaving an entirely sterile Chapter 4 that reads like a statistical printout. Others write their discussion of the literature, policy implications, and theoretical contributions inside Chapter 4, leaving nothing substantive for Chapter 5.
✔ How to Fix It
The distinction is this: Chapter 4 presents and interprets your findings in relation to your research questions and objectives. Chapter 5 discusses your findings in relation to the existing literature, draws conclusions, and makes recommendations. In Chapter 4, you answer the question: “What did your data show?” In Chapter 5, you answer: “What does what your data showed mean — in the context of what other scholars have found, what theory predicts, and what practitioners and policymakers should do about it?” Keep those conversations in their correct chapters. Chapter 4 may briefly note where a finding is consistent with a prior study — but the sustained comparative discussion belongs in Chapter 5.
7. Mistake 6: Losing Alignment With Chapter 1 Objectives
Mistake 6
Common Error
Chapter 4 presents findings that do not correspond one-to-one with the specific objectives stated in Chapter 1
This is the internal consistency failure that examiners find most difficult to overlook. A student states four specific objectives in Chapter 1. Chapter 4 presents findings organised around three themes that roughly correspond to three of those objectives — but the fourth objective is partially addressed across multiple sections, or not addressed at all. Alternatively, Chapter 4 presents findings on a topic that was not among the stated objectives, because the student collected and found interesting data that was not part of the original design.
✔ How to Fix It
Organise Chapter 4 explicitly by objective. Each major sub-section of Chapter 4 should be headed by or clearly labelled with the objective it addresses — for example: “4.3 Findings on Objective Two: To examine the influence of collateral requirements on loan application outcomes.” Every finding presented under that heading must speak directly to that objective. No objective from Chapter 1 should be absent from Chapter 4. No major finding in Chapter 4 should be unconnected to a Chapter 1 objective. If you discover during analysis that an additional finding is genuinely important, the scholarly approach is to discuss it in Chapter 5 as an emergent or incidental finding — not to silently add it to Chapter 4 as if it were part of the original design.
🇰🇪 The Alignment Check Kenyan Examiners Perform
Before approving a dissertation for examination, supervisors and internal examiners at KU, UoN, JKUAT, and MKU routinely perform a simple column-by-column check: they list the specific objectives from Chapter 1, the research questions from Chapter 1, the instrument sections from Chapter 3, and the sub-sections of Chapter 4 — and verify that there is a one-to-one correspondence across all four columns. If any cell in that grid is empty, they send the chapter back. Build this grid yourself before you submit and check every cell.
8. Mistake 7: Weak Qualitative Analysis — Quotation Without Synthesis
Mistake 7
Common Error
Qualitative findings presented as a sequence of interview quotations with minimal analytical commentary between them
The qualitative equivalent of SPSS output dumping is quotation dumping — reproducing large blocks of interview or focus group transcript and providing little or no analytical commentary beyond “Participant 4 stated that…” followed by a block quote, then “Participant 7 said…” followed by another block quote. This approach demonstrates that data was collected, but not that it was analysed.
Thematic analysis — the most widely used qualitative approach in Kenyan university postgraduate research — requires the analyst to identify patterns across multiple data sources, construct themes that represent those patterns, and explain what the themes mean in the context of the research questions. Themes are analytical constructs, not category labels. “Access to Finance” is a category label. “Perceived exclusion from formal credit systems as a driver of informal borrowing among urban youth” is a theme — it makes an analytical claim about what the data shows.
✔ How to Fix It
Present each theme with an analytical paragraph that describes what the theme represents and how it emerged from the data. Use quotations selectively — one or two well-chosen quotations per theme, chosen because they exemplify the pattern most clearly, not because they are the longest or most dramatic. After each quotation, analyse it: explain what it reveals about the participant’s experience, how it reflects or departs from the broader pattern in your data, and what it means for the research question the theme addresses. The ratio of student analysis to participant quotation in a well-written qualitative findings chapter should be at least 60:40.
✍️ Qualitative Presentation — Weak vs. Strong
Weak: “Participant 3 stated: ‘We tried to get a loan from the bank but they wanted us to have title deeds and we don’t own land.’ Participant 7 said: ‘The bank told me my business was too small.’ Participant 11 noted: ‘I don’t understand the forms they give you.'”
Strong: “A central theme emerging from the interview data was the perception of formal financial institutions as structurally inaccessible to respondents’ category of enterprise. This was not simply a matter of eligibility failure, but a deeper experience of institutional mismatch — participants consistently described collateral and documentation requirements as designed for a different economic category than the one they occupy. One participant illustrated this structural gap clearly: ‘We tried to get a loan from the bank but they wanted us to have title deeds and we don’t own land.’ This sentiment was shared by eleven of the fourteen participants interviewed, suggesting that it reflects a systemic rather than individual experience of exclusion. The significance of this theme for the study’s second objective lies in its implication that credit risk assessment frameworks in formal banking institutions are calibrated to asset ownership as a proxy for creditworthiness — a proxy that systematically disadvantages landless micro-enterprise operators regardless of their trading performance.”
9. Mistake 8: Ignoring or Misreporting Reliability and Validity Results
Mistake 8
Common Error
Omitting pilot study reliability results, or reporting a Cronbach’s Alpha value without explaining what it means
In Chapter 3, most students at Kenyan universities commit to testing the reliability of their instruments using Cronbach’s Alpha, with a threshold of ≥0.70 regarded as acceptable. In Chapter 4, this commitment must be honoured — but many students either skip it entirely, report a single coefficient without per-construct breakdowns, or report a value below threshold without acknowledging the problem it creates.
✔ How to Fix It
Present your Cronbach’s Alpha results in a clearly labelled table that shows the coefficient for each construct or variable group in your instrument — not just for the instrument as a whole. A composite alpha of 0.83 may conceal a sub-scale with an alpha of 0.52, which would invalidate the findings for that sub-scale. Report construct-level alphas. State explicitly whether each value meets the ≥0.70 threshold (Nunnally, 1978; George & Mallery, 2003) and, if any do not, explain the implications for how those findings should be interpreted. For content validity, briefly describe the expert review process: how many experts reviewed the instrument, their qualifications, and what revisions were made as a result of their feedback.
Mistake 9
Common Error
Tables without titles or clear headings; figures without axis labels; inconsistent numbering; tables that contain more information than the surrounding text addresses
A table or figure in a dissertation is not decoration and it is not a data dump. It is a communication tool — its entire purpose is to convey a specific piece of information to the reader more efficiently than prose alone could. When a table has no title, or its columns are unlabelled, or it contains twelve variables when the discussion that follows addresses only three, it fails that purpose entirely.
✅ Table and Figure Standards to Follow
- Every table has a title above it — Table 4.3: Descriptive Statistics for Credit Risk Management Variables
- Every figure has a title below it — Figure 4.1: Conceptual Framework Variable Correlation Map
- All columns and rows are labelled in plain English — no SPSS abbreviations left unexplained
- Tables are numbered by chapter: Table 4.1, 4.2, 4.3; Figures 4.1, 4.2
- APA 7th: no vertical lines in tables; horizontal lines only at top, below column headers, and at the bottom
- Table notes appear below the table to explain abbreviations or significance codes
- Tables are referenced in the text before they appear — “as shown in Table 4.3…”
❌ Table and Figure Errors to Avoid
- Tables with no title, or with titles that simply say “Table 4” or “Results”
- SPSS column headers left as-is: “Sig. (2-tailed)”, “Std. Error”, “B” with no explanation
- Pie charts to display more than five categories — they become unreadable
- Bar charts with no axis labels or values displayed
- Tables that appear in the text before they are mentioned in the prose
- Duplicate presentation — same data in both a table and a figure in the same section
- Inconsistent numbering — Table 4.1, Table 4.2, then suddenly Table 5 or Table A
11. Mistake 10: Failing to Link Findings Back to the Literature
Mistake 10
Common Error
Chapter 4 presents findings in isolation, with no connection to the theoretical or empirical literature reviewed in Chapter 2
While the sustained engagement with literature belongs in Chapter 5, Chapter 4 should not be a literature-free zone. Findings presented without any reference to the body of knowledge your study sits within feel unmoored — they tell the reader what your data shows but give no indication of whether this is surprising, confirmatory, contradictory, or novel in the context of prior research. Examiners notice this absence and typically flag it with comments like: “The findings are presented but not contextualised” or “The relationship between your results and the existing literature is not established.”
✔ How to Fix It
Within Chapter 4, after presenting and interpreting each major finding, include one or two sentences that briefly situate it in relation to the literature — without fully developing the comparison (that is Chapter 5’s work). For example: “The finding that collateral requirements are the primary barrier to credit access among micro-enterprises in this study is broadly consistent with emerging evidence from Kenya’s informal financial sector (Ochieng, 2023; Waweru & Ngugi, 2022), although the magnitude of the effect found here — with 78.4% of respondents citing this as the primary barrier — is notably higher than the 61% reported in Ochieng’s Kisumu-based study, suggesting a possibly more pronounced constraint in the Nairobi context.” This single sentence connects your result to the literature, flags a point of comparison, and signals the analytical thread that Chapter 5 will develop more fully.
12. The Chapter 4 Pre-Submission Checklist
Before you submit Chapter 4 to your supervisor, work through this checklist section by section. Every item represents a documented reason why Kenyan university supervisors and examiners send Chapter 4 back for revision.
Response Rate and Data Adequacy
- Response rate: Is the response rate clearly calculated and reported as a percentage? Is it compared to the accepted threshold (≥70% for quantitative studies)? Is any shortfall explained?
- Reliability: Are Cronbach’s Alpha results reported at the construct level — not just as a composite? Does every construct meet the ≥0.70 threshold, or are deviations explained?
- Demographic profile: Is the respondent profile presented in a table? Does it cover the key demographic variables relevant to your study (gender, age, level of education, years of experience, or other applicable categories)?
Quantitative Findings
- Descriptive statistics: Is every table interpreted in prose — not just restated? Does the interpretation explain what the mean or frequency distribution means for the research question it addresses?
- Test selection: Is every inferential test appropriate for the data type and measurement scale of the variables involved? Is normality tested and reported where parametric tests are used?
- Regression or correlation results: Are the key coefficients (R², F, β, p) reported and interpreted in plain language? Is statistical significance correctly distinguished from practical significance?
- Hypothesis testing: If hypotheses were stated in Chapter 1, is each one explicitly accepted or rejected in Chapter 4, with the statistical basis stated?
Qualitative Findings
- Thematic analysis: Are themes analytical constructs — not just category labels? Does the student’s analytical voice dominate over participant quotations?
- Quotation use: Are quotations used selectively — one or two per theme — and followed by interpretive commentary?
- Credibility: Are member-checking, triangulation, or other credibility strategies mentioned and briefly described?
Structural Integrity
- Objective alignment: Does every major sub-section of Chapter 4 correspond to a stated objective from Chapter 1? Is every Chapter 1 objective addressed by at least one finding in Chapter 4?
- Tables and figures: Do all tables have titles above them, correct numbering, and no vertical lines (APA 7th)? Do all figures have labelled axes and titles below them?
- Literature connection: Does each major finding include at least a brief sentence connecting it to the existing literature reviewed in Chapter 2?
- Chapter summary: Does Chapter 4 end with a summary paragraph that recaps the key findings across all objectives and signals the transition to Chapter 5’s discussion?
13. How Tobit Research Consulting Can Help
Chapter 4 is the chapter that requires both technical competence — knowing which tests to run and how to run them — and scholarly judgement: knowing what the results mean, how to communicate them clearly, and how to keep the analytical thread of your research objectives running consistently from the first table to the final paragraph. That combination is exactly where many Kenyan postgraduate students get stuck, and where supervision alone is often not enough to get unstuck quickly.
At Tobit Research Consulting, we work with Masters and PhD students across Kenya at every stage of Chapter 4 — from running and interpreting SPSS, Stata, R, and NVivo analysis, to restructuring findings presentations that have been returned by supervisors, to reviewing complete chapter drafts against the specific rubric criteria used by your institution. We do not conduct analysis on your behalf — we work with you to ensure you understand what your data is telling you and can write it up in the way your examiner expects to read it.
Chapter 4 Data Analysis Support for Kenyan Masters and PhD Students
Tobit Research Consulting provides expert, integrity-focused data analysis and findings presentation support for students at KU, UoN, JKUAT, MKU, Strathmore, Egerton, Moi, Laikipia, Kisii University, and universities across Kenya. Our Chapter 4 services include:
- SPSS, Stata, R, EViews, and NVivo analysis support — running tests and explaining results
- Test selection guidance — choosing the right inferential test for your data type and research question
- Descriptive statistics interpretation — writing analytical prose around your frequency and mean tables
- Regression, correlation, ANOVA, and chi-square result interpretation in full narrative form
- Qualitative thematic analysis — theme construction and findings write-up for interview and focus group data
- Reliability analysis — Cronbach’s Alpha calculation and reporting at construct level
- Chapter 4 structural review — alignment check against Chapter 1 objectives and Chapter 3 methodology
- Table and figure formatting to APA 7th and institutional standards
- Full Chapter 4 draft review and supervisor-ready revision support
- Turnitin similarity checking and plagiarism reduction
- Complete dissertation support from proposal to final submission and viva preparation
Whether you are beginning your data analysis or revising a chapter that has been returned by your supervisor, we are here to help you get it right.
Book a Free Consultation →
📍 Bruce House, 4th Floor, Nairobi CBD, Kenya | Tel: +254 728 430 728 | tobitresearchconsulting.com
This guide is part of Tobit Research Consulting’s Postgraduate Research Skills Series. Key methodological sources informing this guide include: Mugenda, O. M. & Mugenda, A. G. (2003). Research Methods: Quantitative and Qualitative Approaches. Acts Press; Kothari, C. R. (2004). Research Methodology: Methods and Techniques. New Age International; Creswell, J. W. (2014). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. SAGE; Nunnally, J. C. (1978). Psychometric Theory. McGraw-Hill; George, D. & Mallery, P. (2003). SPSS for Windows Step by Step. Allyn & Bacon; Lincoln, Y. S. & Guba, E. G. (1985). Naturalistic Inquiry. SAGE; Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101; APA Publication Manual, 7th Edition (2020); and institutional dissertation guidelines from Kenyatta University, University of Nairobi, JKUAT, and Mount Kenya University.
