Difference-in-Differences (DiD) Analysis
Estimate the effect of a policy or intervention by comparing the change in a treated group with the change in a control group. You get the 2×2 means table, the DiD regression, a two-way fixed effects estimate and an event study that checks for parallel pre-trends.
Updated · By Bibin XavierData 640 × 5
| Storetext · 80 levels | Yearnumeric | Treatednumeric · coded | Salesnumeric | Store_Sizenumeric |
|---|---|---|---|---|
| S01 | 2016 | 1 | 72.7 | 1796 |
| S01 | 2017 | 1 | 67.5 | 1796 |
| S01 | 2018 | 1 | 73.9 | 1796 |
| S01 | 2019 | 1 | 71.1 | 1796 |
| S01 | 2020 | 1 | 78.8 | 1796 |
| S01 | 2021 | 1 | 81.8 | 1796 |
| S01 | 2022 | 1 | 80.6 | 1796 |
| S01 | 2023 | 1 | 86 | 1796 |
Variables and options
Results
640 rows · example dataDifference-in-differences: group means
| Group | Before | After | Change |
|---|---|---|---|
| Treated | 61.355 (n = 160) | 73.467 (n = 160) | 12.112 |
| Control | 57.401 (n = 160) | 63.691 (n = 160) | 6.291 |
| Difference | 3.954 | 9.776 | 5.821 |
DiD regression (Clustered by unit (80 clusters) standard errors)
| Variable | Coefficient | SE | t | p | 95% CI |
|---|---|---|---|---|---|
| Constant | 57.4006 | 1.0975 | 52.299 | < .001 | [55.2160, 59.5852] |
| Treated | 3.9544 | 1.3946 | 2.835 | .006 | [1.1784, 6.7303] |
| Post | 6.2906 | 0.2329 | 27.006 | < .001 | [5.8270, 6.7543] |
| Treated × Post (DiD) | 5.8212 | 0.3203 | 18.172 | < .001 | [5.1836, 6.4589] |
Two-way fixed effects (unit and time fixed effects)
| Estimate | Coefficient | SE | t | p | 95% CI |
|---|---|---|---|---|---|
| Treatment effect (ATT) | 5.8212 | 0.3213 | 18.115 | < .001 | [5.1816, 6.4609] |
Note. Clustered by unit. With treatment starting at the same time for all treated units, TWFE equals the canonical DiD; with staggered adoption it can be biased (Goodman-Bacon, 2021).
Outcome over time by group
Event study: effect relative to the period before treatment
Event-study coefficients
| Relative period | Coefficient | SE | p | 95% CI |
|---|---|---|---|---|
| -4 | -0.055 | 0.753 | .942 | [-1.554, 1.444] |
| -3 | -0.535 | 0.738 | .470 | [-2.003, 0.933] |
| -2 | -0.042 | 0.736 | .954 | [-1.508, 1.423] |
| 0 | 4.445 | 0.682 | < .001 | [3.088, 5.802] |
| 1 | 4.120 | 0.691 | < .001 | [2.744, 5.496] |
| 2 | 6.888 | 0.696 | < .001 | [5.503, 8.272] |
| 3 | 7.200 | 0.652 | < .001 | [5.901, 8.499] |
Note. Joint test that all pre-treatment coefficients are zero: F(3, 79) = 0.249, p = .861. Reference period: −1.
Interpretation
After treatment, the outcome of the treated group changed by 5.82 more than that of the control group (95% CI [5.18, 6.46]), which is statistically significant. The pre-treatment coefficients are jointly insignificant, consistent with parallel trends.
How to cite this tool
Using these results in a thesis, journal article, report or class assignment? Please cite this free tool so others can find it. It keeps the site free.
APA 7
Xavier, B. (2026). Difference-in-Differences (DiD) Analysis [Web application]. FreeStatisticalTools.com. https://freestatisticaltools.com/difference-in-differences-calculator/
BibTeX
@software{xavier_differenceindifferences_2026,
author = {Xavier, Bibin},
title = {Difference-in-Differences (DiD) Analysis},
publisher = {FreeStatisticalTools.com},
year = {2026},
url = {https://freestatisticaltools.com/difference-in-differences-calculator/},
note = {Web application}
}Suggested sentence for your methods section: “The analysis was carried out with the free Difference-in-Differences (DiD) Analysis tool on FreeStatisticalTools.com (Xavier, 2026).”
What is difference-in-differences?
Difference-in-differences estimates a causal effect by comparing how an outcome changed for a group exposed to a treatment with how it changed for an unexposed group. The comparison removes stable differences between groups and common shocks over time.
When to use difference-in-differences
Use DiD to evaluate a policy, programme or event that affected some units and not others, when you observe both groups before and after.
Assumptions
- Parallel trends: without treatment, both groups would have followed the same path (check the event study).
- No anticipation and no spill-overs to the control group.
- Treatment starts at the same time for all treated units in this tool; staggered adoption needs dedicated estimators.
How to set up your data
Long format: one row per unit and period with the outcome, a 0/1 treated-group indicator, the time variable and optionally a unit identifier.
How to use this tool
- Enter your data. Paste columns from Excel, Google Sheets or SPSS with variable names in the first row, or upload a CSV or Excel file. The example data shows the expected layout.
- Choose the variables. Select the outcome, treatment group indicator, value for the treated group and time variable.
- Set the options. Fill in first treated period.
- Read the results. Results update automatically. Read the plain-English interpretation, then use Copy table to paste APA-style tables into Word and Copy for the results sentence.
How to report difference-in-differences in APA style
Report the DiD estimate with clustered standard errors, the group means, and the event-study plot with the pre-trend test.
Frequently asked questions
What is the parallel trends assumption?
That without the treatment, the treated and control groups would have followed the same trend. Pre-treatment event-study coefficients close to zero support it.
How do I interpret the interaction coefficient?
The Treated × Post coefficient is the DiD estimate: the extra change in the outcome for the treated group after treatment.
What if units are treated at different times?
Staggered adoption can bias two-way fixed effects. Use estimators designed for it, such as Callaway and Sant’Anna (2021).
Related tools
Browse all categories
Data cleaning and preparation · Describe and explore · Compare means · Nonparametric tests · Correlation and regression · Bayesian statistics · Categorical data · Reliability and agreement · Multivariate analysis · Structural equation modeling · Econometrics and time series · Panel data and causal inference · Marketing and decision research · Survival, clinical and meta-analysis · Study planning and calculators · PLS-SEM