User 136 · Research · Groups · Tf038 Proof Ttest
Executive Summary

Executive Summary

Whether the two group means differ and how large the effect is

N Total
120
N Group 1
60
N Group 2
60
Mean Group 1
71.614
Mean Group 2
77.476
T-Statistic
-2.6632
Degrees of Freedom
117.29
P-Value
0.0088
Mean Difference
-5.862
Cohen d
-0.4862
The difference in means between control (71.61) and treatment (77.48) is statistically significant (p = 0.0088 < alpha = 0.05). The raw mean difference is -5.862 (95% CI: [-10.221, -1.503]). Cohen's d = -0.486, indicating a small effect size. The evidence supports a real difference between the group means.
Interpretation

The difference in means between control (71.61) and treatment (77.48) is statistically significant (p = 0.0088 < alpha = 0.05). The raw mean difference is -5.862 (95% CI: [-10.221, -1.503]). Cohen's d = -0.486, indicating a small effect size. The evidence supports a real difference between the group means.

Data Table

Descriptive Statistics by Group

Sample sizes, means, standard deviations, and 95% confidence intervals per group

GroupNMeanSDSECI LowerCI Upper
control6071.6112.521.61668.3874.85
treatment6077.4811.581.49574.4880.47
Interpretation

control has n = 60 observations with mean = 71.61 (SD = 12.52); treatment has n = 60 with mean = 77.48 (SD = 11.58). Standard errors are 1.616 and 1.495 respectively. 95% confidence intervals for the group means are [68.38, 74.85] and [74.48, 80.47].

Visualization

Group Means with Confidence Intervals

95% CI for each group mean — non-overlapping CIs suggest a significant difference

Interpretation

Group means with 95% confidence intervals: control = 71.61 [68.38, 74.85], treatment = 77.48 [74.48, 80.47]. The confidence intervals overlap, which is consistent with a non-significant difference. The t-test p-value = 0.0088 provides the formal significance test.

Visualization

Value Distribution by Group

Box plots comparing spread, median, and outliers between the two groups

Interpretation

Box plots show the spread and central tendency of each group's outcome values. control has median = 70.16 (IQR = 19); treatment has median = 77.12 (IQR = 15.32). The relative position of the boxes indicates whether one group consistently scores higher, while box width and whisker length reflect within-group variability and potential outliers.

Visualization

T-Test Results and Effect Size

Key statistics from Welch's t-test including t-statistic, p-value, mean difference, and Cohen's d

Interpretation

Welch's t-test: t(117.3) = -2.663, p = 0.0088. The mean difference (control minus treatment) is -5.862. Cohen's d = -0.486, which is a small effect. The result is statistically significant (p = 0.0088 < alpha = 0.05).

Visualization

Distribution Shape by Group

Within-group value histograms for visual normality assessment

Interpretation

Histograms show the value distribution within each group, allowing visual assessment of the normality assumption required by the t-test. control appears approximately normal (Shapiro-Wilk p = 0.6778); treatment appears approximately normal (Shapiro-Wilk p = 0.8126). The t-test is robust to mild non-normality, especially when both groups have n ≥ 30.

Data Table

Assumption Test Results

Shapiro-Wilk normality tests per group and Levene's test for variance equality

TestStatisticP ValueConclusion
Shapiro-Wilk: control0.98510.6778Normality supported
Shapiro-Wilk: treatment0.98780.8126Normality supported
Levene's Test (Equal Variances)0.97340.3258Equal variances supported
Interpretation

Shapiro-Wilk tests for control (p = 0.6778) and treatment (p = 0.8126) assess normality. Levene's test for equal variances gives p = 0.3258 — Equal variances supported. Both groups are consistent with normality, supporting parametric t-test validity.

Your data has more stories to tell. Run any analysis on your own data — 60+ validated R modules, interactive reports, AI insights, and PDF export. 2,000 free credits on signup.
Try Free — No Signup Sign Up Free

Report an Issue

Tell us what's wrong. You'll get a free re-run of this analysis so you can try again with different parameters. If the re-run still doesn't meet your expectations, we'll refund your credits.

Want to run this analysis on your own data? Upload CSV — Free Analysis See Pricing