Mastering Stata for Data Analysis: Übungsmappe 2 Tutorial 2025/26

Introduction: Why Data Analysis Skills Matter in 2026

In today's data-driven world, the ability to analyze survey data using tools like Stata is more valuable than ever. Whether you're tracking public opinion on privacy, predicting election outcomes, or understanding social trends, these skills are essential. This tutorial guides you through key tasks from the Angewandte Datenanalyse Übungsmappe 2. Teil (Winter 2025/26), focusing on variable preparation, cross-tabulation, and index building using the ALLBUS 2016 dataset. By the end, you'll be ready to tackle your assignment with confidence.

Getting Started: Setting Up Your Stata Do-File

Your assignment requires submitting a .do file and a PDF. Start by downloading the template from Moodle and renaming it as Uebungsmappe_02_YourLastName.do. Open it in Stata and update the file path to your local folder containing the ALLBUS dataset (ZA5251_v1-1-0.dta). Remember to write your full name in the designated line. Organize your code under the provided section headers, and always test that your do-file runs without errors.

Task 1: Variable Preparation and Univariate Description

Restricting the Sample by Age

You need to focus on respondents who were at least 36 and at most 85 years old at the time of the 2016 survey. This ensures they experienced part of their youth before German reunification. Use the age variable (or generate it from birth year) and apply keep if age >= 36 & age <= 85. Check the age range with summarize age.

Creating a Dichotomous Variable for Strong Opinion on Video Surveillance

Variable J011_1 measures attitudes toward video surveillance in public spaces. To create vid_stark, recode values 1 (strongly in favor) and 5 (strongly against) to 1, and all others (2,3,4, -1, etc.) to 0. Use:

recode J011_1 (1=1) (5=1) (else=0), gen(vid_stark)
label variable vid_stark "Strong opinion on video surveillance"
label define vid_stark_lbl 1 "Yes" 0 "No"
label values vid_stark vid_stark_lbl

Verify with tab vid_stark J011_1, missing to ensure correct mapping.

Creating a Dummy for Youth in East Germany

Variable dg03 indicates where the respondent lived at age 15. Recode it into jugendost where 1 = East Germany and 0 = West Germany. For example:

recode dg03 (1/2=1) (3/4=0) (else=.), gen(jugendost)
label variable jugendost "Jugend in Ostdeutschland"
label define ost_lbl 1 "ja" 0 "nein"
label values jugendost ost_lbl

Check with tab jugendost dg03, missing.

Describing vid_stark with One Number

To describe the variable meaningfully, report the proportion of respondents with a strong opinion. For instance, if 40% have a strong opinion, you can say: "40% of the sample hold a strong opinion on video surveillance (either strongly in favor or strongly against)." Use tab vid_stark to get the percentage.

Task 2: Cross-Tabulation and Hypothesis Testing

Your hypothesis: People who grew up in East Germany are more likely to have a strong opinion on video surveillance. Use the prepared variables:

tab jugendost vid_stark, row chi2 V

This produces a cross-tabulation with row percentages, the Chi-squared test, and Cramer's V. Interpret the table: compare the percentage of strong opinion among East vs. West. For example, if 50% of East youth have a strong opinion vs. 35% of West, that supports the hypothesis. The p-value from Chi² tells if the association is statistically significant (p < 0.05). Cramer's V indicates effect size (0.1 small, 0.3 medium, 0.5 large). Write a clear interpretation in your PDF.

Task 3: Building an Additive Index for Law & Order Attitudes

You are asked to create an index from items J013_1, J013_2, J014_1, J014_2, J014_3. First, check their coding: are they all on the same scale? If some are reversed (e.g., higher values mean less support), you must reverse-code them. Use recode to align directions. For example, if an item ranges 1-7 and 7 means strong support, but another item uses 1 for support, reverse the latter. After alignment, create the index as the sum (or mean) of non-missing items. Use:

alpha J013_1 J013_2 J014_1 J014_2 J014_3, gen(law_order_index) // mean index
egen law_order_index = rowmean(J013_1 J013_2 J014_1 J014_2 J014_3) // alternative

Check internal consistency with Cronbach's alpha (should be >0.7). Label the new variable and describe its distribution.

Trend Connection: Privacy Debates in 2026

As of May 2026, debates about government surveillance and data privacy are hotter than ever. With the rise of AI-powered facial recognition in public spaces, understanding public opinion on video surveillance is crucial. Your analysis of ALLBUS data contributes to this conversation, showing how regional socialization shapes attitudes. This mirrors real-world discussions in Germany and beyond.

Final Tips for Your Übungsmappe Submission

Ensure your do-file runs without errors. Remove any broken code lines.
In the PDF, write full German sentences interpreting each result. Do not copy the entire do-file.
Include the signed declaration of originality on the last page.
Name files as Uebungsmappe_02_YourLastName.pdf and .do.
Double-check that you used the correct ALLBUS dataset (ZA5251 v1-1-0).

By following these steps, you'll demonstrate mastery of Stata for data analysis. Good luck with your assignment!