AI-Therapy | Statistics for Psychologists

About AI-Therapy

AI-Therapy creates online fully automated programs using the latest evidence-based treatments, such as cognitive behavioural therapy. The social anxiety program has been tested with a randomized controlled trial. To find out more visit:

These results were generate by Comparing two sets of data on September 26, 2023 at 8:45 PM (UTC).

Results

Data set overview

If you would like to share these results, please use the following link:

https://www.ai-therapy.com/psychology-statistics/results/20230926204556824

Data set name	Python Test Correctness
Data set notes	(not provided)

Data set statistics

Sample name	Number of samples	Mean	Standard error of the mean	Standard deviation	Median
Copilot	360	0.608	0.026	0.489	1.000
ChatGPT	360	0.714	0.024	0.453	1.000

Test results

Number of samples	N = 720
Wilcoxon rank-sum	W = 122940.000 z-score = -2.451 Asymptotic significance (2-tailed) p = 0.014 Based on a p-value threshold of 0.01, there is not a statistically siginificant difference between 'Copilot' and 'ChatGPT'. Effect size r = -0.091
Mann-Whitney	U = 57960.000 z-score = -2.990 Asymptotic significance (2-tailed) p = 0.003 Based on a significance level of 0.01, there is a statistically siginificant difference between 'Copilot' and 'ChatGPT'. Effect size r = -0.111

Visualization

The graph above is a bihistogram that allows you to compare the distribution of Copilot (top) with the distribution of ChatGPT (bottom).