How to Use SPSS for Data Analysis in a Dissertation or Thesis

Using SPSS for dissertation data analysis is a multi-stage process that extends well beyond clicking buttons in the Analyze menu. Examiners evaluate your entire analytical pipeline: how you prepared data, whether you checked assumptions, whether test selection matched your design, and whether you reported results completely and correctly. This guide provides a practical SPSS workflow for thesis and dissertation students—from raw survey export through hypothesis testing to APA-formatted results paragraphs ready for your data analysis chapter.

Stage 1: Import and organise data

Import from Excel (.xlsx), CSV, or Google Forms export via File → Open or File → Import. Check for duplicate respondent IDs. Assign clear variable names (max 64 characters, no spaces—use underscores). Open Variable View and set Type (Numeric, String, Date), Label (full description), Values (code labels for categories), and Measure (Nominal, Ordinal, Scale).

Stage 2: Clean and recode data

• Set missing values: blank cells or coded values like 99.
• Reverse-score negatively worded Likert items: Transform → Recode.
• Compute scale totals: Transform → Compute Variable.
• Filter or split file for subgroup analyses when needed.
• Identify outliers with box plots or z-scores; document exclusion criteria.

Stage 3: Descriptive statistics

Run Analyze → Descriptive Statistics → Frequencies for categorical variables (gender, department). Run Descriptives or Explore for continuous variables (means, SDs, min, max). Present in demographic profile table in your thesis. Cross-tabulations (Analyze → Descriptive → Crosstabs) show relationships between categorical variables.

Stage 4: Reliability analysis

For multi-item scales, run Analyze → Scale → Reliability Analysis (Cronbach's alpha). Alpha ≥ 0.70 is generally acceptable; ≥ 0.80 is good. Report alpha for each scale in methodology. Remove items that reduce alpha if theoretically justified.

Stage 5: Check assumptions before inferential tests

• Normality: Analyze → Descriptive → Explore with normality plots, or Shapiro-Wilk.
• Homogeneity of variance: Levene's test in t-test and ANOVA output.
• Linearity: scatterplot for correlation and regression.
• Multicollinearity: VIF in regression output (VIF < 10).
• Expected frequencies: chi-square output checks.

Stage 6: Run hypothesis tests

Match each hypothesis to one SPSS procedure:

• Two groups, continuous DV: Analyze → Compare Means → Independent-Samples T Test.
• Paired measurements: Analyze → Compare Means → Paired-Samples T Test.
• 3+ groups: Analyze → Compare Means → One-Way ANOVA.
• Two continuous variables: Analyze → Correlate → Bivariate.
• Prediction: Analyze → Regression → Linear.
• Two categorical variables: Analyze → Descriptive → Crosstabs → Chi-square.

Stage 7: Read SPSS output correctly

Focus on the correct table: t-test uses 'Independent Samples Test' (equal variances row). ANOVA uses 'ANOVA' table F-statistic. Regression uses 'Coefficients' table for β and p. Correlation uses Pearson Correlation matrix. Always note exact p-value, test statistic, and df.

Stage 8: Calculate effect sizes

SPSS does not always print effect sizes automatically. Compute Cohen's d for t-tests, η² for ANOVA, r² for regression using formulas or Effect Size calculators. Effect sizes demonstrate practical significance alongside p-values.

Stage 9: Create thesis-ready tables and charts

Copy tables from Output Viewer; paste as editable objects in Word. Format to APA: no vertical lines, clear headings, notes for abbreviations. Charts: Graphs → Chart Builder for histograms, bar charts, and scatterplots. Label axes with units.

Stage 10: Write APA results paragraphs

Example: 'An independent samples t-test compared satisfaction scores between online and offline learners. Online learners (M = 4.21, SD = 0.58) reported significantly higher satisfaction than offline learners (M = 3.87, SD = 0.62), t(178) = 3.45, p = .001, d = 0.51.' One paragraph per hypothesis test.

Dissertation SPSS analysis help

Our SPSS data analysis service handles the full pipeline for your thesis—data cleaning, assumption checks, hypothesis testing, and APA results writing—while you retain ownership of your research.