To interpret your SPSS output you read each results table from the top down: confirm the test that ran, check the test statistic and its degrees of freedom, then read the significance value (the Sig. column, which is the p-value) and finally the effect size. SPSS hands you several boxes per analysis, and only a few numbers in each one actually answer your research question. The skill is knowing which cells to quote in your dissertation and which to ignore.
Read the boxes in the order SPSS prints them
SPSS does not give you one answer; it gives you a stack of tables, and they are meant to be read in sequence. The early boxes describe your sample (case counts, means, standard deviations), the middle boxes test your assumptions, and the final box reports the result you will write up. If you jump straight to the last table you risk quoting a p-value from a test whose assumptions were already violated two boxes earlier. Reading top to bottom is the habit that keeps your results chapter defensible, and it mirrors the structure you will see when you move from descriptive to inferential statistics.
The four numbers that carry your result
Almost every inferential output reduces to four numbers. The test statistic (a t, an F, or a chi-square) tells you how large the effect is in test units. The degrees of freedom reflect your sample size and design. The Sig. value is the probability of seeing a result this extreme if there were no real effect, and SPSS prints it to three decimals (a value of .000 means less than .001, never exactly zero). The effect size, such as Cohen's d or eta squared, tells you whether a significant result is also a meaningful one. Those four map directly onto the parts you will label in your write-up.
Reading a regression table without getting lost
Regression output is where most students stall, because SPSS prints three boxes at once. The Model Summary gives R and R squared, which is the share of variance in your outcome explained by your predictors. The ANOVA box tests whether the model as a whole beats an empty model; its Sig. value should be small before you trust anything else. The Coefficients box is the one you quote most: each predictor's B is its slope, the Beta is the standardised version that lets you compare predictors, and each row's own Sig. tells you whether that predictor matters once the others are held constant. Before you trust any of it, confirm the model meets the assumptions behind linear regression.
Turning the cells into examiner-ready prose
An output table is not a result until you have written it as a sentence. Pull the four numbers, state the direction of the effect in plain words, and attach the effect size so the reader knows whether the finding is large or trivial. A significant p-value with a tiny effect size is a real but unimportant result, and saying so shows methodological maturity. The formatting conventions for those numbers, including italics, decimals, and the no-leading-zero rule for probabilities, are set out in how to report statistics in APA style, and the surrounding narrative belongs in your dissertation results chapter.
| SPSS box | What you read from it |
|---|---|
| Descriptives | Means, standard deviations, case counts |
| Model Summary | R and R squared (variance explained) |
| ANOVA | Whether the overall model is significant |
| Coefficients | Each predictor's slope, Beta, and Sig. |
A short worked reading
Suppose you ran a regression predicting exam score from study hours and prior grade. The Model Summary shows R squared of .42, so your two predictors explain forty-two percent of the variance in scores. The ANOVA box reports a Sig. of .001, so the model as a whole is sound. In the Coefficients box, study hours has a B of 1.8 with a Sig. of .003, while prior grade has a Sig. of .21. You would write that study hours significantly predicted exam score whereas prior grade did not, once both were in the model together. That single paragraph is the payoff of reading the output in order and quoting only the cells that matter.