Survey Research and Analysis How-To

Analysis Studio: The Complete How-To Guide

How to use Analysis Studio from first load to finished result: setup, survey designs, licensing, and every one of the twenty modules step by step.

How to Weight a Survey, Step by Step

The full weighting workflow in order: base weights, nonresponse adjustment, raking and post-stratification, trimming, and the diagnostics that prove the work.

What Is a Design Effect?

The design effect prices complex samples and weighting in one number. How it is computed, what drives it, and how it converts a nominal sample into an effective one.

How Many Responses Do You Need?

Sample size for surveys and studies, worked out in the open: margins of error, group comparisons, design effects, and response rates, each with the arithmetic shown.

Statistical Power, Explained with Worked Numbers

What power is, what moves it, and what an underpowered study costs, with the numbers worked: power curves, detectable effects, and the four levers.

Effect Sizes: Reporting How Much, Not Just Whether

Cohen d, odds ratios, Cramér V, eta squared, and hazard ratios, each with its interval and a worked figure, plus the bands and their limits.

Which Statistical Test Should You Use?

A decision guide across the Analysis Studio roster: by outcome type, grouping structure, pairing, and survey design, to the procedure that answers the question honestly.

Reading A/B Test Results

The difference is the headline, the interval is the story, and the sample ratio mismatch check decides whether either can be trusted. A worked test, read properly.

Propensity Scores for Nonprobability Samples

How pseudo-weighting an opt-in sample against a reference probability sample works, what it corrects, what it cannot correct, and how to report it.

Survey Trend Analysis Across Waves

Testing change over time in repeated cross-sectional surveys: per-wave estimates, polynomial trend contrasts on real time values, and two-wave change tests done honestly.

Cronbach's Alpha: Interpretation and Reporting

Reading alpha honestly: the bands, the item-count effect, alpha if item deleted, why high alpha is not unidimensionality, and a reporting paragraph that holds up.

Reading a Kaplan-Meier Curve

Censoring marks, the number at risk, median survival, and the log-rank test, taught on one small curve computed step by step.

Analysis Studio Module Guides

A guided tour of every Analysis Studio module: what it computes, what it needs, and how to read its results.

Analysis Studio

Advanced analytics, right in your browser. Weighting, regression, survival analysis, and more, on your own data.

Every Monday

Statistics for Investors

The statistics behind investing, worked on real data: averages that mislead, risk measures and their limits, and what backtests can and cannot show.

Wednesdays, alternating with Survey Research

Valuation

Company valuation with the arithmetic shown: discounted cash flow, multiples, and the assumptions that move the answer.

Fridays, alternating with Maine by the Numbers

Small Area Estimation

Estimates for counties, towns, and other small places, where direct survey numbers run too thin to stand on their own.

Fridays, alternating with Small Area Estimation

Maine by the Numbers

Maine in data: population, work, housing, and health, county by county, from public sources.