Chapter 3 Statistical Analysis Plan Help — Dissertation Proposal Support

What Belongs in a Chapter 3 Statistical Analysis Plan?

Chapter 3 is written before you collect data, and its statistical section has one job: convince your committee that you know exactly what you’ll do with the data once you have it. That means naming the specific test for each research question, not describing your topic again, and not deferring the decision with “the data will be analysed using SPSS.”

A complete statistical plan specifies:

• Your overall research design and why it fits your research questions
• How each variable will be operationalised and measured
• The exact statistical test planned for each research question or hypothesis
• An a priori power analysis justifying your target sample size
• The validity and reliability evidence for any instrument or scale you’re using

Justifying Your Research Design

Before naming tests, your design needs a clear label and justification: correlational (examining relationships without manipulation), experimental (random assignment to conditions), or quasi-experimental (group comparison without random assignment, common in applied and educational settings). Committees expect this choice tied explicitly to your research questions: a design chosen because it’s “easier to run” rather than because it fits the question is a common weakness reviewers flag.

Naming the Test for Each Research Question

This is the section most Chapter 3 drafts get wrong: research questions listed without a corresponding, named statistical test attached to each one. For every research question or hypothesis, state:

• The variables involved and their measurement level (nominal, ordinal, continuous)
• The specific test that matches that combination (see the SPSS statistical test guide for the full decision logic)
• The assumptions that test requires, and how you’ll check them once data is collected

“Research Question 1 will be analysed using SPSS” is not a complete answer. “Research Question 1 will be tested using a one-way ANOVA, with Levene’s test checked for homogeneity of variance and Tukey post-hoc comparisons if the omnibus test is significant” is.

A Priori Power Analysis and Sample Size Justification

Committees generally expect a priori power analysis: calculated before data collection, not after. This requires four interlocking numbers: your alpha level (conventionally .05), your desired power (conventionally .80), your expected effect size (from prior literature or a conservative default if none exists), and the resulting required sample size. G*Power is the most common companion tool used alongside SPSS for this calculation. See the full power analysis and sample size guide for the step-by-step walkthrough. This section of Chapter 3 is often the single most scrutinised paragraph in a proposal defence.

Validity and Reliability Plan for Your Instruments

If your study uses a survey or scale, Chapter 3 needs to address its psychometric properties before data collection: cite prior validity/reliability evidence for an established instrument, or describe your plan to assess reliability (typically Cronbach’s alpha) once your own data is collected. Committees want to see this addressed proactively, not left until Chapter 4 raises a problem.

Common Chapter 3 Mistakes Committees Flag

• Naming a general analysis approach (“quantitative analysis will be performed”) instead of a specific test per research question
• Missing or clearly post-hoc power analysis
• No stated plan for handling assumption violations if they occur
• Instrument validity/reliability not addressed at all

Once your plan is approved and data is collected, the next stage is reporting what actually happened. See the full Chapter 4 results guide for that transition, or the broader SPSS dissertation and thesis statistics help overview for how all three chapters fit together.