SPSS: From data preparation to reporting

What SPSS is and what this guide is

IBM SPSS Statistics is statistical software for managing data, running analyses and producing tables and charts. It is not a research design and it does not choose a defensible analysis on your behalf. ResearchKit is independent of IBM: this guide explains an educational workflow and links to IBM's own documentation; ResearchKit does not provide an installer, license or technical support for SPSS.

The Lab reads shared ResearchProjectDraft context when a workspace is open. It does not import an SPSS file, upload data, edit the project or save new SPSS state. Menu instructions and feature availability below are checked against IBM SPSS Statistics 32 documentation; dialog labels and licensing may vary by release, operating system and edition. Confirm a path in the documentation for the version you use.

1. 1Research question and design
2. 2Variables and data preparation
3. 3Descriptive checks
4. 4Select a test with the Statistical Test Finder
5. 5Run the mapped procedure in SPSS
6. 6Review assumptions with the Assumption Checker
7. 7Calculate a suitable effect size
8. 8Read output and interpret the reported values
9. 9Write and preserve the analysis report

• IBM SPSS Statistics documentation and release versions (external link)
• IBM SPSS Statistics 32 installation and help (external link)

Recognize the editor, output and syntax

The Data Editor has two complementary views: Data View contains case values; Variable View contains the variable definitions and labels applied to those values. The Output Viewer collects procedure tables, notes and charts. The Syntax Editor holds commands that can be inspected, saved and rerun. Menu dialogs can paste their corresponding syntax, giving you a record of how an analysis was specified.

Save the data, syntax and output as distinct project artifacts. Exporting output for a report is separate from keeping the editable SPSS output file; use the export options present in your installed version and verify the exported tables after conversion. IBM menu labels and available file formats can change between releases.

• IBM SPSS Statistics version 32 documentation (external link)

Start with the research question and variables

Write down the question, target population, design and unit of analysis first. Mark the outcome, predictor or grouping factor, any covariate, and whether the question compares groups, examines association, predicts an outcome, or compares a single sample with a value. This question-first sequence follows established research-methods teaching (Field, 2018; Howell, 2013). The Statistical Test Finder owns the deterministic test-selection reasoning; the SPSS Lab only maps an already selected method to a procedure.

• Use the Statistical Test Finder
• Review and document variables
• Build the broader analysis plan

Create, code and check the data file

In the Data Editor, Data View organizes cases as rows and variables as columns. Variable View records metadata for each column: a concise unique variable name, a human-readable label, value labels for category codes, missing-value definitions, and a measurement level. SPSS uses nominal, ordinal and scale as its measurement labels; scale combines interval and ratio for many procedures. A software label never repairs an invalid measurement or study design.

Coding is defined by the researcher and documented in the codebook; no code is universal. For one questionnaire, gender might use 1, 2 and 3 for locally defined responses; education might use ordered codes from secondary study to doctoral study; a Likert item might use 1 to 5 from strongly disagree to strongly agree. State every label, preserve meaningful missing categories, and do not treat arbitrary category numbers as measurements with equal intervals.

• Keep an untouched source copy; record the imported file, date and any transformations.
• Check IDs and duplicate cases, valid ranges, category labels, impossible combinations, missing-value codes and each analysis's eligible N.
• Inspect frequency tables for categorical variables and distributions/plots for quantitative variables before choosing summaries or inferential methods.
• Reverse-code only the specified items using the documented scale endpoints; verify the resulting minimum and maximum.
• Compute a total or mean scale score only when the item definitions and missing-item rule justify it; keep the source items and record the scoring rule.
• For regression, encode categorical predictors with explicit contrasts/dummy variables; the numeric codes alone do not model categories correctly.
• Record exclusions, recodes, computed variables and outlier decisions before interpreting outcomes.

• IBM SPSS Statistics Data Preparation documentation (external link)

Describe before testing

Frequencies can show counts, percentages and, when requested, descriptive statistics and charts. Descriptives summarizes numeric variables with N, mean, minimum, maximum, standard deviation and variance. Explore supports distributions, percentiles and graphical checks. Match the statistic to the measurement: mode/frequencies for categories, median and quartiles for ordered or skewed values, and mean/SD for quantitative values where those summaries are informative.

• Analyze → Descriptive Statistics → Frequencies: use for category counts/percentages and optionally mode, median, quartiles, charts or other statistics.
• Analyze → Descriptive Statistics → Descriptives: use for numeric summaries such as mean, SD, variance, range, minimum and maximum.
• Analyze → Descriptive Statistics → Explore: use to inspect distributions, percentiles and potential outliers alongside plots.
• An outlier or non-normality flag prompts inspection and methodological reasoning; it does not automatically invalidate a method or require a nonparametric replacement.

• IBM Frequencies documentation (external link)
• IBM Descriptives documentation (external link)

Choose the analysis before opening its dialog

The Statistical Test Finder asks only for information relevant to the question and shows candidate methods, uncertainty, reasons and assumptions. After deciding with that tool and your research context, use the SPSS Lab to locate the corresponding procedure. The one-variable goodness-of-fit chi-square and two-variable chi-square test of independence have different inputs and answer different questions.

Do not switch methods solely because a preliminary normality test crosses a threshold. Consider what parameter the research question concerns, the actual distribution and outliers, sample size, independence, robustness, and whether the alternative answers the same question. The Assumption Checker describes checks and alternatives; the researcher remains responsible for the decision.

• Open the Statistical Test Finder
• Open the SPSS Research Lab

Follow the SPSS procedure for the selected method

These procedure cards reuse ResearchKit's analysis-method IDs, test profiles and assumption knowledge. Menu labels and selected options below were checked against IBM SPSS Statistics 32 documentation. Most referenced procedures require Statistics Base; verify your licensed edition. A single menu path does not establish that the method suits the design.

Check assumptions with the existing Assumption Checker

The Statistical Assumption Checker owns the assumption definitions, design-aware checks, thresholds and remedies. Use it to prepare a checklist, then inspect the relevant data and model diagnostics in SPSS. Independence comes from the sampling/design; a menu option cannot repair dependent observations. For a paired t-test, examine the differences. For regression and GLM models, inspect residuals and the model structure.

• Normality is not a universal pass/fail switch; consider the relevant residuals or paired differences, plots, sample size and robustness.
• Homogeneity tests such as Levene's are evidence to consider, not an automatic command to change tests.
• Outlier rules should be justified by data quality and design; do not remove valid cases solely to obtain a preferred p-value.
• If assumptions are questionable, consult the method's alternatives and supervisor; record the decision and sensitivity checks.

• Plan checks with the Statistical Assumption Checker

Effect size, statistical significance and practical importance differ

A p-value is evidence relative to a specified null model; an effect size describes a chosen measure of magnitude; practical importance depends on consequences, context, measurement and uncertainty. None implies the others (Wasserstein & Lazar, 2016). Conventional magnitude thresholds are context-bound guides, not universal laws (Cohen, 1988). For t-tests, name the standardizer; for paired data, d_z and d_av use different denominators (Hedges, 1981; Lakens, 2013). For ANOVA, distinguish eta squared from partial eta squared. For contingency tables, use a measure suited to the table design.

• Mean

  x̄ = Σx ÷ n

  Where

  x̄

  the sample mean

  Σx

  the sum of all the values

  n

  the number of values

  Add every value and divide by how many there are.

• Sample standard deviation

  s = √[ Σ(x − x̄)² ÷ (n − 1) ]

  Where

  s

  the sample standard deviation

  x

  each value

  x̄

  the sample mean

  n

  the number of values

  The typical distance of values from the mean. Dividing by n − 1 rather than n makes it a better estimate of the population's spread.

• The idea behind every t-test

  t = difference ÷ standard error of the difference

  Where

  difference

  the difference being tested, such as between two group means

  standard error

  how much that difference would vary from sample to sample by chance

  A t value compares the difference found with the variation expected by chance. The further t is from zero, the less the difference looks like chance alone.

• One-sample t

  t = (x̄ − μ₀) ÷ (s ÷ √n), with n − 1 degrees of freedom

  Where

  x̄

  the sample mean

  μ₀

  the value the mean is compared with, stated in advance

  s

  the sample standard deviation

  n

  the number of participants

  How far the sample mean lies from the stated value, measured in standard errors.

• Paired-samples t

  t = d̄ ÷ (s_d ÷ √n), with n − 1 degrees of freedom

  Where

  d̄

  the mean of the differences between each participant's two measurements

  s_d

  the standard deviation of those differences

  n

  the number of pairs

  A paired t-test is a one-sample t-test on the differences, comparing their mean with zero.

• Expected count in a cross-tabulation

  E = (row total × column total) ÷ N

  Where

  E

  the count expected in a cell if the two variables were unrelated

  N

  the total number of participants

  If the variables were unrelated, each cell would hold its row's share of its column's total.

• Chi-square statistic

  χ² = Σ (O − E)² ÷ E

  Where

  χ²

  the chi-square statistic

  O

  the observed count in a cell or category

  E

  the expected count in the same cell or category

  Σ

  add the result for every cell or category

  The larger the gaps between observed and expected counts, relative to the expected counts, the larger χ².

• Simple linear regression

  Y = β₀ + β₁X + ε

  Where

  Y

  the outcome

  X

  the predictor

  β₀

  the intercept, the predicted outcome when X is zero

  β₁

  the slope, the change in the predicted outcome for each one-unit increase in X

  ε

  the error, what the line doesn't explain for each participant

  The outcome is modelled as a straight line in the predictor, plus what the line leaves unexplained.

• Multiple linear regression

  Y = β₀ + β₁X₁ + β₂X₂ + … + βₖXₖ + ε

  Where

  Y

  the outcome

  X₁ … Xₖ

  the k predictors

  β₀

  the intercept

  β₁ … βₖ

  each predictor's slope, holding the other predictors constant

  ε

  the error

  Each slope estimates a predictor's contribution while the other predictors are held constant.

• The F ratio in ANOVA

  F = variance between groups ÷ variance within groups

  Where

  variance between groups (mean square between)

  how far the group means spread around the overall mean

  variance within groups (mean square within)

  how far participants spread around their own group's mean

  When the group means differ by much more than participants differ within groups, F is large.

The Effect Size Calculator shows inputs, formulas and substitutions. It does not calculate significance or declare practical importance. IBM SPSS Statistics 32 can estimate effect sizes in its one-sample and independent/paired t-test procedures and one-way ANOVA when the relevant effect-size option is selected; available standardizers and tables differ by procedure.

• Calculate an effect size
• Read the ASA statement on p-values (external link)

Find the value, then interpret and report it

Read the named output table and row for the effect you planned. Record the correct test row (for example, the variance row justified by the assumptions for an independent t-test), degrees of freedom, p-value, estimate, confidence interval and effect size. For GLM, distinguish the factor, interaction, covariate and error rows. For regression, distinguish model fit from an individual predictor's coefficient. For crosstabs, inspect expected counts and the correct association statistic.

The Results Interpretation Assistant validates and interprets the numbers you enter, linking them to the project where possible. It does not read an output file or recompute raw data. An APA-style result should identify the analysis, statistic and degrees of freedom where applicable, p, effect-size measure and relevant interval; accompany it with the design-specific substantive explanation and limitations.

• Interpret output values
• Build an APA-style results table

Work through synthetic examples

The datasets below are fixed and invented for teaching. They show research questions, variable roles, data structure, candidate method and summaries. They are not empirical findings, are too small to support strong inferential conclusions, and do not contain fabricated SPSS p-values or completed APA results.

• Employee satisfaction across two bank types: identify the outcome, independent grouping factor and independent-samples t-test structure.
• The same participants before and after training: identify pairs and why the analysis concerns within-person differences.
• Study hours and exam score: decide whether the question is symmetric association (correlation) or prediction (regression).
• Exam scores across three teaching groups: explain why one overall ANOVA precedes justified pairwise follow-up.
• Observed customer service-channel counts: distinguish goodness-of-fit against stated shares from independence in a two-variable crosstab.

Keep a reproducible analysis record

SPSS syntax is a record of commands that can be saved and rerun. For example, `FREQUENCIES VARIABLES=department.` is a short syntax command for a frequency table; inspect the pasted command and adapt variable names to your file. Preserve syntax with the input data, output, codebook and decision log. A reproducible record explains sample exclusions, coding, transformations, model specification, assumption checks, output values and reporting choices. Syntax is not a substitute for understanding the analysis.

1. 1Research question and design
2. 2Variables and data preparation
3. 3Descriptive checks
4. 4Select a test with the Statistical Test Finder
5. 5Run the mapped procedure in SPSS
6. 6Review assumptions with the Assumption Checker
7. 7Calculate a suitable effect size
8. 8Read output and interpret the reported values
9. 9Write and preserve the analysis report

Common mistakes and limits

• Choosing a test because it is familiar or because the software menu makes it easy, rather than because the question and design support it.
• Treating independent groups as paired, or paired measurements as independent.
• Confusing one-variable goodness-of-fit chi-square with a two-variable test of independence.
• Ignoring data coding, duplicate cases, missingness, distributions, expected counts or residual diagnostics.
• Changing the analysis after seeing results without a prespecified methodological reason.
• Reporting only p, treating p < .05 as practical importance, or treating a large effect-size label as statistical significance.
• Copying output tables without identifying which row answers the research question or acknowledging multiple comparisons.
• Treating correlation or regression as proof of causation.

ResearchKit supports methodological reasoning; it cannot inspect your SPSS dataset, verify its provenance, certify assumptions, choose among all valid models, or replace supervisor/statistician judgment. When your design includes clustering, survey weights, substantial missingness, complex sampling, repeated-factor designs, or methods outside the supported catalogue, seek specialist guidance.

References and official SPSS documentation

Operational menu paths and procedure descriptions link directly to IBM SPSS Statistics 32 documentation. Methodological references support the statistical explanations and effect-size distinctions.

• Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates.
• Field, A. (2018). Discovering statistics using IBM SPSS statistics (5th ed.). SAGE.
• Hedges, L. V. (1981). Distribution theory for Glass's estimator of effect size and related estimators. Journal of Educational Statistics, 6(2), 107–128. https://doi.org/10.3102/10769986006002107 (external link)
• Howell, D. C. (2013). Statistical methods for psychology (8th ed.). Wadsworth, Cengage Learning.
• Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4, Article 863. https://doi.org/10.3389/fpsyg.2013.00863 (external link)
• Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: Context, process, and purpose. The American Statistician, 70(2), 129–133. https://doi.org/10.1080/00031305.2016.1154108 (external link)

• IBM SPSS Statistics documentation index (external link)
• IBM SPSS Statistics 32 documentation (external link)

Frequently asked questions

Related guides

• How to choose a statistical test

  A practical, step-by-step guide to matching a research question, variables and study design with statistical tests, assumptions and interpretation.

  Read guide

• How to report statistics in APA Style

  How to write statistical results in APA Style: formatting t, F, r, χ² and p, decimal places and leading zeros, effect sizes and confidence intervals, example sentences for common tests, and when to use a table.

  Read guide

• What a p-value tells you

  What a p-value measures, what it doesn't, the most common misinterpretations, why statistical significance isn't importance, and how to report and read p-values alongside effect sizes and confidence intervals.

  Read guide

Related tools

• Statistical Test Finder

  Helps you identify statistical tests that may suit your research question and data, and explains why each may fit.

  Open tool

• Effect Size Calculator

  Calculate selected standardized effects from reported summary statistics, with formulas and assumptions shown.

  Open tool

• SPSS Research Lab

  Connects a research question and selected analysis to IBM SPSS Statistics procedures, output tables, assumptions and reporting tools.

  Open tool

• Variables Builder

  Brings your variables together from your project, and helps you define, operationalise and measure each one.

  Open tool

• Data Analysis Recommender

  Recommends statistical analyses from your variables, hypotheses, design and sample, and explains why each one suits your project.

  Open tool

• Results Interpretation Assistant

  Explains the statistical results you already have: what they mean, whether they are significant, and how they bear on your hypotheses, objectives and research question.

  Open tool

• Statistical Assumption Checker

  Explains which assumptions to check before each statistical analysis, how to check them, the usual thresholds, and what to do if they fail.

  Open tool

• New

  Research Table Builder

  Builds publication-ready research tables in APA, IEEE, Harvard, Chicago or MLA style, from your data, your results or your project, ready to paste into Word.

  Open tool