SPSS (Statistical Package for the Social Sciences) remains one of the most widely used statistical software packages in academic research. This comprehensive tutorial will guide you through using SPSS for your dissertation, from initial data entry to advanced statistical analyses.
Getting Started with SPSS
When you open SPSS, you'll see two main views: Data View (where you see your actual data in spreadsheet format) and Variable View (where you define the properties of each variable). Understanding both views is essential for effective data management.
Key interface elements: The menu bar provides access to all statistical procedures. The toolbar offers shortcuts to common operations. The output window displays all results. The syntax editor (optional) allows you to write and save commands for reproducibility.
Pro tip: Always save your work in two formats—the .sav data file AND a syntax file (.sps) documenting your analyses. This ensures reproducibility and makes revisions easier.
Data Entry and Import
Manual data entry: In Variable View, first define your variables with appropriate names (no spaces, max 64 characters), types (numeric, string, date), and value labels. Then switch to Data View to enter your data.
Importing from Excel: Go to File → Open → Data, select your Excel file, and SPSS will guide you through the import process. Ensure your Excel file has variable names in the first row and data starting from row 2.
Variable types to know: Nominal (categories with no order, e.g., gender), Ordinal (ordered categories, e.g., education level), Scale/Continuous (numerical data with meaningful intervals, e.g., age, test scores).
Setting up your variables correctly:
``` Name: participant_id (unique identifier) Type: Numeric Measure: Nominal Name: age Type: Numeric Measure: Scale Name: satisfaction Type: Numeric (1-5) Value Labels: 1='Very Dissatisfied', 2='Dissatisfied', 3='Neutral', 4='Satisfied', 5='Very Satisfied' Measure: Ordinal ```
Data Cleaning and Preparation
Before analysis, clean your data thoroughly. Common issues include missing values, outliers, and data entry errors.
Identifying missing values: Analyze → Descriptive Statistics → Frequencies. Check for unexpected values or high missing data rates. Consider whether data is missing completely at random (MCAR), missing at random (MAR), or missing not at random (MNAR).
Handling missing data options: Listwise deletion (exclude cases with any missing data), pairwise deletion (use all available data for each analysis), mean substitution (replace with variable mean), or multiple imputation (most robust for significant missing data).
Detecting outliers: Use Analyze → Descriptive Statistics → Explore. Check boxplots for extreme values. Calculate z-scores (values beyond ±3.29 are typically outliers). Decide whether to remove, transform, or retain outliers based on your methodology.
Recoding variables: Transform → Recode into Different Variables. Useful for creating categories from continuous variables or collapsing response categories.
Computing new variables: Transform → Compute Variable. Create composite scores, reverse-code items, or calculate derived measures.
Descriptive Statistics
Descriptive statistics summarize your data and should be reported for all key variables in your dissertation.
For continuous variables (Analyze → Descriptive Statistics → Descriptives or Explore): Report mean, standard deviation, range, skewness, and kurtosis. Check for normality using histograms and the Shapiro-Wilk test.
For categorical variables (Analyze → Descriptive Statistics → Frequencies): Report frequencies and percentages. Use bar charts or pie charts for visualization.
Creating publication-ready tables: Right-click output tables → Copy Special → select 'Objects' to paste into Word with formatting intact.
Example output interpretation: ``` Descriptive Statistics N Mean Std. Deviation Job Satisfaction 245 3.72 0.89 Work Experience 245 12.45 6.32 Age 245 38.21 9.87 ```
Inferential Statistics
Inferential statistics help you draw conclusions about your population from your sample data.
Choosing the right test: Your choice depends on your research question, number of groups/variables, and whether your data meets parametric assumptions (normality, homogeneity of variance).
Independent samples t-test (Analyze → Compare Means → Independent Samples T-Test): Compares means between two unrelated groups. Example: comparing test scores between treatment and control groups. Report t-value, degrees of freedom, p-value, and effect size (Cohen's d).
Paired samples t-test (Analyze → Compare Means → Paired Samples T-Test): Compares means within the same group at two time points. Example: pre-test vs post-test scores. Report the same statistics plus the correlation between measures.
One-way ANOVA (Analyze → Compare Means → One-Way ANOVA): Compares means across three or more groups. Include post-hoc tests (Tukey or Bonferroni) to identify which groups differ. Report F-value, degrees of freedom, p-value, and eta-squared.
Chi-square test (Analyze → Descriptive Statistics → Crosstabs): Tests association between two categorical variables. Report χ², degrees of freedom, p-value, and Cramér's V for effect size.
Correlation (Analyze → Correlate → Bivariate): Measures relationship strength between continuous variables. Use Pearson for normally distributed data, Spearman for ordinal or non-normal data. Report r-value, p-value, and interpret strength (weak: .1-.3, moderate: .3-.5, strong: .5+).
Regression Analysis
Regression analysis predicts outcomes and identifies significant predictors—essential for many dissertation research questions.
Simple linear regression (Analyze → Regression → Linear): One predictor, one outcome. Reports R², adjusted R², F-test, unstandardized (B) and standardized (β) coefficients, and p-values.
Multiple regression (Analyze → Regression → Linear with multiple predictors): Multiple predictors, one outcome. Check for multicollinearity using VIF (values >10 indicate problems). Methods: Enter (all predictors simultaneously), Stepwise (automatic selection), Hierarchical (theory-driven blocks).
Interpretation example: ``` Model Summary: R² = .342, Adjusted R² = .331 The model explains 33.1% of variance in job satisfaction. Coefficients: B SE β t Sig. (Constant) 1.245 .312 3.99 .000 Work-life balance .456 .078 .412 5.85 .000 Salary satisfaction .289 .065 .298 4.45 .000 Team support .178 .071 .167 2.51 .013 ```
Logistic regression (Analyze → Regression → Binary Logistic): For binary outcome variables. Reports odds ratios, confidence intervals, and Nagelkerke R².
Exporting Results for Your Dissertation
Professional presentation of results is crucial. SPSS offers several export options:
Export tables: Right-click table → Export → choose Word, Excel, or PDF format. For APA formatting, you may need to adjust tables in Word.
Export charts: Right-click chart → Copy or Export. For high-quality graphics, export as PNG or EMF.
Creating syntax files: File → New → Syntax. Record your analyses as commands for reproducibility. This is essential for dissertation committees who may want to verify your analyses.
APA reporting format examples:
- t-test: "An independent samples t-test revealed a significant difference in satisfaction between groups, t(243) = 2.84, p = .005, d = 0.42."
- ANOVA: "A one-way ANOVA showed significant differences across conditions, F(2, 242) = 5.67, p = .004, η² = .045."
- Regression: "Work-life balance was a significant predictor of job satisfaction (β = .41, p < .001), controlling for salary and team support."
Common SPSS Mistakes to Avoid
1. Incorrect variable measurement levels: Always verify that SPSS correctly identifies your variables as Nominal, Ordinal, or Scale.
2. Ignoring assumption checks: Run normality tests and check homogeneity of variance before parametric tests.
3. Not saving syntax: Always document your analyses for reproducibility.
4. Misinterpreting significance: Statistical significance doesn't equal practical importance—always report effect sizes.
5. Using default settings blindly: Customize analyses based on your research design and data characteristics.
Next Steps
This tutorial covers SPSS fundamentals for dissertation research. For advanced techniques like factor analysis, structural equation modeling, or multilevel modeling, consider consulting with a statistician or taking advanced statistics courses. Remember, the best statistical analysis is one that directly addresses your research questions while respecting your data's characteristics and your field's reporting conventions.
