Dissertation Data Analysis Guide UAE: SPSS, Excel & NVivo (2026)

Use your programme’s requirements, then select the analysis

There is no single “UAE Chapter 4” length, order or software rule. A master’s thesis, doctoral dissertation, capstone and journal manuscript can have different chapter structures. For example, American University of Sharjah directs researchers to its Office of Graduate Studies for current thesis formatting and submission guidance. Your supervisor-approved proposal and institution’s current handbook govern your own document.

Before opening SPSS, Excel or NVivo, write one row for each research question: the variable or material being studied, how it was collected, the planned analysis, the output that will answer the question, and the limitation you will disclose. This links methods to results and reveals gaps while they can still be discussed with your supervisor.

What SPSS, Excel and NVivo are useful for

Mixed-methods work may use more than one tool. The research design, type of evidence and analysis question come first; software follows. See Labeeb’s tool comparison for a narrower choice guide.

From a survey file to a defensible result

1. Preserve the raw file. Keep a read-only original, a codebook and a dated record of exclusions, recodes and missing-data decisions.
2. Describe the actual sample. Report invitations, responses, usable cases and important missingness. A calculated sample-size target is not proof that the achieved sample represents the population.
3. Choose the test for the question and measurement. Group comparisons, relationships and prediction models answer different questions. If a planned test is unsuitable, document why you changed it rather than trying many tests until one returns a small p-value.
4. Check assumptions for the selected method. For linear regression, IBM’s SPSS documentation describes data and model assumptions and available diagnostics. A blanket instruction to test normality, homogeneity and multicollinearity before every statistical method is incorrect; the checks depend on the model.
5. Report the result, then interpret cautiously. Provide the relevant sample size, estimate, uncertainty and test information required by your discipline or style. Explain what the estimate means for the research question without turning association into causation.

APA’s Journal Article Reporting Standards provide separate quantitative, qualitative and mixed-methods reporting guidance. They can help with transparent reporting, but they do not replace your university’s thesis template.

From transcripts to an auditable interpretation

1. Document the sampling, consent and permitted use of interview or document material; remove unnecessary personal identifiers before external review.
2. Define how material was prepared and which version was analysed. Keep a clear record of corrections to transcription and translation.
3. Code the material in a way that fits the chosen methodology. Software helps organise excerpts; it does not create valid themes by itself.
4. Keep analytic notes showing why codes were combined, separated or rejected. Retain contradictory examples where they matter.
5. Write findings with enough contextual evidence to support each claim, while protecting participants according to the approved ethics protocol.

Do not claim that every UAE programme requires NVivo, a fixed number of themes or a specific chapter length. The APA qualitative and mixed-methods reporting standards can be a useful transparency checklist when your programme permits them.

Write what an SPSS coefficient supports

Imagine a hypothetical survey in which the fitted regression coefficient for a predictor is 0.72. That number alone does not mean “72% improvement” or “strong causal impact.” To interpret it, the researcher must identify the outcome and predictor units, the model, the adjustment variables, the estimate’s uncertainty and the study design. A careful sentence could state that the fitted outcome was higher by an estimated 0.72 units for each one-unit difference in the predictor within this analysed sample, subject to the stated model assumptions. These figures are illustrative, not Labeeb or client research results.

Likewise, a theme label in NVivo is not a finding unless the researcher can show how the source material and analysis support it. Make the path from question to data to conclusion visible.

A practical results-chapter check

• Every major table, figure or theme answers a stated research question.
• Labels, units, sample sizes, missing-data notes and source periods are consistent.
• Reported values agree with the preserved analysis output; do not retype or round until meaning changes.
• Methods and changes from the proposal are explained honestly.
• The discussion separates a result from a possible explanation, limitation or recommendation.
• Formatting follows the current programme requirements; raw software output is not automatically wrong, but a reader should be able to understand it without navigating a full output dump.

If you need help understanding or presenting your own analysis, Labeeb’s data-analysis support can be scoped around the dataset, approved methods, output review and a walkthrough. You remain responsible for your research choices, interpretation and submission.

Methods and reporting references

Sources reviewed 14 October 2026: AUS thesis submission guidance; IBM SPSS linear-regression documentation; APA Journal Article Reporting Standards. These references illustrate sound method and reporting checks; the approved protocol and current institution instructions take precedence for an individual project.