DIY Explorer
Explore raw data. Produce frequencies, correlations, and crosstabs to understand relationships among key variables during this pivotal time in US history.
How to Use This Tool
Browse weighted frequencies, correlations, and crosstabs from the November 1963 Harris–Newsweek survey
How to Use This Tool
Browse weighted frequencies, correlations, and crosstabs from the November 1963 Harris–Newsweek survey
1. Select Variables
Check the variables you want to explore. Hover the info icon on any card for the original question wording. The six predictors from the published approval model are pre-selected as a default — deselect any you don’t need or add others from the full list of 128 survey items. Use the search box and category filters to find specific variables quickly.
2. Choose Analysis Type
Choose from three analysis modes. Frequencies shows the weighted distribution of responses for each selected variable. Correlations computes pairwise Pearson correlations across all selected variables — useful for spotting which items move together. Crosstabs breaks one variable’s distribution down by the categories of another; select exactly two variables for this mode.
3. Run Analysis
Click Run Analysis to compute and display results. All calculations use the survey’s design weights, so estimates reflect the population rather than the raw sample. The results section appears below the form — scroll down to see tables, charts, and summary statistics. A notification at the top of the page confirms when results are ready.
4. Pin Runs to Compare
Click Pin this run in the results header to save a snapshot of the current output. Pinned runs collect in a drawer at the bottom of the screen — expand it to review and compare previous results without losing your current view. This is useful for comparing frequencies across different variable subsets or for tracking how a correlation changes when you swap variables in and out.
5. Filter and Search
Use the Search variables box to find items by keyword or question code (e.g., “q31” or “Khrushchev”). The category filter chips let you narrow the variable list by topic — Political, Policy, Evaluation, Demographics. Active filters are shown as chips above the variable list; click the × on any chip to clear it. The selected-variable count updates in real time.
6. Iterate
Change your variable selection or analysis type and re-run as many times as you like. The Model Builder (next page) lets you take variables you’ve identified here and use them as predictors in a logistic regression — so the Explorer is a natural first step for deciding which variables are worth modeling.