SPSS for Beginners: A Step-by-Step Guide to Data Analysis

IBM SPSS Statistics is the most widely used statistics program in graduate theses across the social and health sciences. Its menu-driven interface lets you run analyses without writing code — but that same ease makes it just as easy to run the wrong test in a few clicks. Getting correct results depends less on knowing the software than on knowing which test answers which question.

1.Step 1: Define your variables (Variable View)

SPSS has two main screens: Data View, where values are entered, and Variable View, where variables are defined. Before entering data, complete these fields for each variable:

• Name: a short name without spaces (e.g. age, gender, item1).
• Label: the full description that appears in output.
• Values: codes for categorical variables (1 = Female, 2 = Male).
• Missing: the code used for missing data (e.g. 99).
• Measure: the level of measurement — Nominal, Ordinal or Scale.

2.Step 2: Clean your data

Run Analyze > Descriptive Statistics > Frequencies to check the minimum and maximum of every variable. A value of 55 on an item scored 1–5 is a data entry error. Check the amount and pattern of missing data. Reverse-score items with Transform > Recode into Different Variables, and compute total scores with Transform > Compute Variable.

3.Step 3: Descriptive statistics

Report counts and percentages for categorical variables and means, standard deviations, medians and ranges for continuous variables. These form the first table in your results chapter.

4.Step 4: Check normality

Parametric tests such as t-tests, ANOVA and Pearson correlation assume approximately normal data. Use Analyze > Descriptive Statistics > Explore. Shapiro–Wilk is typically reported for smaller samples and Kolmogorov–Smirnov for larger ones, but in large samples these tests flag even trivial deviations, so also inspect skewness and kurtosis, histograms and Q–Q plots. Skewness and kurtosis within ±1.5 or ±2 is a commonly used criterion.

5.Step 5: Choose the right test

• Two independent groups: independent-samples t-test / Mann–Whitney U
• Three or more independent groups: one-way ANOVA / Kruskal–Wallis
• Same group measured twice: paired-samples t-test / Wilcoxon signed-rank
• Two categorical variables: chi-square / Fisher’s exact test
• Two continuous variables: Pearson / Spearman correlation
• Predicting an outcome from one or more variables: linear or logistic regression

6.Step 6: Reliability analysis

If you used a scale, calculate internal consistency in your own sample with Analyze > Scale > Reliability Analysis. A Cronbach’s alpha of .70 or higher is generally considered acceptable. Calculate it separately for each subscale.

7.Step 7: Read and report the output

SPSS produces many tables; do not paste them all into your thesis. Present only the statistics that answer your research questions, rebuilt as APA-style tables. For example: “Female participants reported higher anxiety (M = 45.2, SD = 8.1) than male participants (M = 40.6, SD = 7.9), t(198) = 4.05, p < .001, d = 0.57.”

8.Common beginner mistakes

• Running parametric tests without checking assumptions
• Skipping post-hoc tests after a significant ANOVA
• Ignoring the expected-count assumption in chi-square tests
• Reporting p-values without effect sizes
• Trying different tests until one is significant