You ran an independent-samples t-test in SPSS. The output viewer now shows four tables, two footnotes, and a box plot you did not ask for. Your thesis needs one sentence and three numbers. The gap between what SPSS gives you and what your results section actually requires trips up students constantly. This guide walks through the output for the four most common tests and shows you exactly which rows to read, which tables to skip, and how to turn each result into an APA-7 formatted sentence.
The Independent-Samples T-Test
When you run Analyze > Compare Means > Independent-Samples T Test in SPSS, the output produces two tables. The first is Group Statistics. It shows the mean (M), standard deviation (SD), sample size (N), and standard error for each group. You need the mean, SD, and N from this table. Write them down or copy them directly into your results paragraph.
The second table is Independent Samples Test, and this is where students get lost because it packs two rows into a single table. The top row is labeled "Equal variances assumed" and the bottom row is "Equal variances not assumed." To decide which row to read, look at the left side of the same table: Levene's Test for Equality of Variances shows an F value and a Sig. column. If that Sig. value is greater than .05, the variances are roughly equal and you read the top row. If Sig. is .05 or below, the assumption of equal variances is violated and you read the bottom row (which uses a Welch correction and adjusted degrees of freedom).
From whichever row applies, you need four values: t (the test statistic), df (degrees of freedom), Sig. (2-tailed) which is your p-value, and Mean Difference. SPSS does not compute Cohen's d for you; calculate it by dividing the mean difference by the pooled standard deviation, or let a tool like Academic Stats Agent handle it automatically.
Suppose your Group Statistics table shows: Intervention group M = 74.2, SD = 8.6, N = 42; Control group M = 68.9, SD = 9.1, N = 38. Levene's Sig. = .614, so you read the "Equal variances assumed" row: t = 2.67, df = 78, Sig. (2-tailed) = .009. The APA sentence reads:
Students in the intervention group scored significantly higher (M = 74.2, SD = 8.6) than those in the control group (M = 68.9, SD = 9.1), t(78) = 2.67, p = .009, d = 0.60.
That is the complete report. No screenshot of the SPSS table. No footnotes. One sentence with means, standard deviations, the test statistic, degrees of freedom, exact p-value, and effect size.
The Paired-Samples T-Test
A paired-samples t-test compares two measurements from the same participants, such as pre-test and post-test scores. SPSS outputs three tables for this analysis: Paired Samples Statistics, Paired Samples Correlations, and Paired Samples Test.
Start with Paired Samples Statistics. This gives you the mean and standard deviation for each measurement. If you are reporting a pre/post design, these are the numbers you describe as "before" and "after." The Paired Samples Correlations table shows the correlation between the two measurements. You rarely report this directly in a thesis, but a very low correlation (below .20) may signal that the paired design was weak, and some committee members will ask about it.
The table that matters most is Paired Samples Test. It reports the mean difference between the two conditions, the standard deviation of that difference, the standard error, a 95% confidence interval for the difference, the t statistic, df, and Sig. (2-tailed). Read t, df, and Sig. from this row. There is only one row, so there is no ambiguity about which to pick.
Example: 45 nursing students completed an anxiety questionnaire before and after a simulation exercise. Paired Samples Statistics shows pre-test M = 36.8, SD = 7.2 and post-test M = 31.4, SD = 6.9. The Paired Samples Test table reports t(44) = 4.12, Sig. = .000. In APA, you never write p = .000 because the actual value is not zero; SPSS rounds anything below .0005 to .000. Write p < .001 instead. For the effect size, divide the mean difference (5.4) by the standard deviation of the differences (which SPSS shows in the same table). If that SD is 8.7, then d = 5.4 / 8.7 = 0.62.
Anxiety scores decreased significantly from pre-test (M = 36.8, SD = 7.2) to post-test (M = 31.4, SD = 6.9), t(44) = 4.12, p < .001, d = 0.62.
One-Way ANOVA
An ANOVA compares means across three or more groups. Run it through Analyze > Compare Means > One-Way ANOVA, and SPSS produces at least three tables: Descriptives, Test of Homogeneity of Variances, and the ANOVA table. If you requested post hoc tests, a fourth table appears.
The Descriptives table lists the mean, standard deviation, sample size, standard error, and 95% confidence interval for each group. Report the mean and SD for every group in your text or in a table within the thesis. The Test of Homogeneity of Variances (Levene's statistic) is diagnostic. It checks whether variances are equal across groups. You do not report this table directly in most theses. If its Sig. is below .05, mention that the homogeneity assumption was violated and consider using the Welch ANOVA or Games-Howell post hoc test instead of Tukey. If Sig. is above .05, you can proceed without comment. (For a deeper comparison of when to choose ANOVA over simpler tests, see our t-test vs. ANOVA guide.)
The ANOVA table is one row with four numbers you need: the Between Groups row gives you the F statistic, df (which actually consists of two values: df between groups and df within groups), and Sig. Eta-squared (η²) is not included in the default output. Calculate it by dividing the Between Groups Sum of Squares by the Total Sum of Squares, both of which appear in the same table. For example, if Between Groups SS = 482.3 and Total SS = 3,210.5, then η² = 482.3 / 3,210.5 = .15.
Suppose you are comparing satisfaction scores across three hospital departments (n = 30 per group). The ANOVA table shows F(2, 87) = 7.34, Sig. = .001, and you calculated η² = .14. The APA sentence:
There was a significant difference in satisfaction scores across the three departments, F(2, 87) = 7.34, p = .001, η² = .14.
That sentence tells the reader there is a difference somewhere, but not where. This is where the Post Hoc table comes in. Tukey HSD or Bonferroni pairwise comparisons list every group-to-group contrast with its mean difference, standard error, Sig., and confidence interval. Report only the significant pairs (and note which pairs were not significant). For example: "Post hoc comparisons using Tukey HSD indicated that the Emergency department (M = 3.92) scored significantly lower than both Cardiology (M = 4.48, p = .003) and Pediatrics (M = 4.41, p = .006). The difference between Cardiology and Pediatrics was not significant (p = .938)."
Chi-Square Test of Independence
Chi-square tests whether two categorical variables are associated. Run it through Analyze > Descriptive Statistics > Crosstabs, check "Chi-square" under Statistics, and request "Expected counts" under Cells. SPSS produces a Crosstabulation table and a Chi-Square Tests table.
The Crosstabulation table shows observed and expected frequencies for every combination of categories. You typically present this as a table in your thesis. Make sure all expected cell counts are 5 or above; if any cell has an expected count below 5, the chi-square approximation is unreliable and SPSS will flag this in a footnote. With a 2 × 2 table and any expected count below 5, report Fisher's Exact Test instead of Pearson Chi-Square.
The Chi-Square Tests table usually lists several variants: Pearson Chi-Square, Continuity Correction, Likelihood Ratio, Fisher's Exact Test, and Linear-by-Linear Association. For most theses, report the Pearson Chi-Square row. Read three values from it: Value (the χ² statistic), df, and Asymptotic Significance (2-sided), which is your p-value. For effect size, Cramér's V appears in a separate Symmetric Measures table if you checked "Phi and Cramér's V" under Statistics. For a 2 × 2 table, Phi and Cramér's V are identical.
Example: you surveyed 200 university students about whether they used tutoring services (yes/no) and whether they passed the final exam (pass/fail). The Chi-Square Tests table shows Pearson χ² = 6.84, df = 1, Asymptotic Sig. = .009. Cramér's V = .185. The APA sentence:
There was a significant association between tutoring use and exam outcome, χ²(1, N = 200) = 6.84, p = .009, V = .19.
Skip the Continuity Correction row (it applies only to 2 × 2 tables and is overly conservative). Skip the Likelihood Ratio row unless your supervisor specifically requests it. Skip Linear-by-Linear Association unless your variables are ordinal and you are testing a trend.
Which Tables to Skip (and Why)
SPSS generates more output than any thesis needs. Knowing what to ignore saves time and prevents the common mistake of pasting screenshots of every table into the appendix. The table below summarizes which SPSS output tables to report and which to skip for each test.
| Test | Report These Tables | Skip These Tables |
|---|---|---|
| Independent t-test | Group Statistics, Independent Samples Test (correct row) | Levene's is diagnostic only; read it to pick your row, do not report it as a standalone result |
| Paired t-test | Paired Samples Statistics, Paired Samples Test | Paired Samples Correlations (report only if theoretically relevant) |
| One-way ANOVA | Descriptives, ANOVA, Post Hoc (if F is significant) | Test of Homogeneity of Variances (diagnostic; mention only if violated) |
| Chi-square | Crosstabulation, Pearson Chi-Square row, Symmetric Measures | Continuity Correction, Likelihood Ratio, Linear-by-Linear Association |
A good rule of thumb: diagnostic tables tell you whether your test assumptions are met. You read them silently and act on them (by switching to a different test or adjusting your procedure). You do not paste them into your results chapter. Reporting tables are the ones that answer your research question. Those go into the thesis. If you are unsure about the meaning of a specific p-value in your output, our guide to interpreting p-values covers the logic in detail.
Formatting Tips That Prevent Revision Requests
After working on over 300 student projects, we see the same formatting errors trigger supervisor feedback round after round. The fixes are small but they add up.
First, never paste a screenshot of SPSS output into your thesis. Committees expect you to extract the relevant numbers and present them in APA format within the text. A screenshot signals that the student does not know which numbers matter. If your supervisor asks for the raw output, include it in an appendix, not inline.
Second, always report exact p-values. SPSS displays "Sig. = .000" when the p-value rounds below .0005. Write p < .001 in your thesis. For any value .001 or above, report the exact figure: p = .034, not p < .05. Remove the leading zero from p-values (because they cannot exceed 1.00), but keep the leading zero on other statistics like Cohen's d = 0.72 or η² = 0.14.
Third, italicize all statistical symbols. That means M, SD, t, F, p, N, d, r, χ², and η². Numbers, parentheses, and equal signs are not italicized. This is spelled out in APA 7th edition Table 6.5, and reviewers notice when it is wrong.
Fourth, always include an effect size. SPSS does not produce Cohen's d, eta-squared, or Cramér's V by default for most tests. You either calculate them by hand, request them through the SPSS options dialogs (e.g., "Estimates of effect size" is available for some newer procedures), or run your analysis through Academic Stats Agent, which includes effect sizes in every output automatically.
Key takeaway: For each SPSS test, you need three things in your thesis: (1) descriptive statistics from the first table (means, SDs, or frequencies), (2) the test statistic, degrees of freedom, and exact p-value from the main test table, and (3) an effect size that you usually have to compute yourself. Everything else in the SPSS output is either diagnostic or redundant. Read it, act on it if needed, but do not paste it into your results section.