Calculator Inputs
Example Data Table
| Scenario | Treatment Mean | Control Mean | Control SD | Treatment SD | Treatment N | Control N | Glass Delta |
|---|---|---|---|---|---|---|---|
| Study A | 12.4 | 9.1 | 4.2 | 5.0 | 40 | 38 | 0.7857 |
| Study B | 18.0 | 15.6 | 3.5 | 4.1 | 52 | 50 | 0.6857 |
| Study C | 7.5 | 8.4 | 2.8 | 2.9 | 30 | 29 | -0.3214 |
Use these rows to test the calculator, compare positive and negative effects, and check how control-group variability changes the final standardized difference.
Formula Used
Glass Delta measures the standardized difference between treatment and control means by using only the control group standard deviation.
Glass Delta = (Treatment Mean − Control Mean) ÷ Control Standard Deviation
Δ = (Mt − Mc) / SDc
Large-sample variance ≈ (nt + nc) / (ntnc) + Δ2 / [2(nt + nc)]
- Treatment Mean: average score for the intervention or target group.
- Control Mean: average score for the baseline group.
- Control Standard Deviation: spread of scores in the control group.
- Sample Sizes: counts used for the standard error and interval estimate.
How to Use This Calculator
- Enter the treatment group mean and the control group mean.
- Add the control group standard deviation carefully, because it drives the denominator.
- Provide both sample sizes to estimate standard error and confidence interval.
- Optionally enter the treatment standard deviation for a side-by-side Cohen's d comparison.
- Select a confidence level and adjust the direction label if needed.
- Press Calculate Glass Delta to show the result above the form.
- Use the CSV or PDF buttons to save the output for reporting.
Why Use Glass Delta?
Glass Delta is helpful when group variances differ meaningfully. Because it standardizes with the control deviation alone, the effect size stays anchored to the baseline group instead of mixing variability from both groups.
That makes it useful for experiments, evaluations, and studies where the treatment changes variability as well as the mean. It is often reported alongside Cohen's d or Hedges' g for context.
FAQs
1. What does Glass Delta measure?
It measures the difference between two group means after scaling that difference by the control group standard deviation.
2. When is Glass Delta better than Cohen's d?
It is often preferred when the treatment and control groups have noticeably different standard deviations, especially when the treatment may alter variability.
3. Can the result be negative?
Yes. A negative value means the treatment mean is below the control mean under the selected subtraction direction.
4. Why does the control standard deviation matter so much?
It is the denominator of the statistic. Smaller control variability produces a larger standardized effect for the same mean difference.
5. Does this calculator support confidence intervals?
Yes. It provides a large-sample interval using the selected confidence level and an approximate standard error formula.
6. What if I only know the means and control deviation?
You can still calculate Glass Delta. The treatment standard deviation is optional and only affects the comparison with Cohen's d.
7. Is this calculator suitable for teaching and reporting?
Yes. It combines computation, interpretation, export tools, and example data, which makes it useful for coursework and practical reporting.