Why do I have to paste every score and not just the average?

Because telling whether the difference is real requires knowing how much scores vary within each group. Two groups with the same average can have very similar or very scattered scores, and that changes the result.

What is Welch’s t-test?

It’s a way to compare two averages that works even when the groups have different sizes or different spread. That makes it a good fit for survey data.

It’s the probability of seeing a difference this large or larger if the two groups actually had the same average. If it’s below 0.05, the difference is called significant at 95%.

How do I get each person’s score?

With a satisfaction survey in each group. In SurvIQ you can create one for free from a template, share it by link, WhatsApp or QR code, and people respond without creating an account.

All tools

Compare two averages calculator

At one branch the average satisfaction is 4.3 and at another 4.1. Is that a real difference? This calculator tells you. It works for 1 to 5 scores, 1 to 10 scores or any scale, across two branches, two months or two products. Paste each person’s score for group A in one box and for group B in the other.

Group A

Name (optional)

Paste this group’s scores

One per line or separated by commas. For example, the column of a 1 to 5 question

Group B

Name (optional)

Paste this group’s scores

One per line or separated by commas. For example, the column of a 1 to 5 question

The calculation runs in your browser. We don’t store what you type.

How it’s calculated

For each group, the calculator finds the average, the number of responses (n) and the sample standard deviation (s), which measures how spread out the scores are.

The standard error of the difference is the square root of (sA² / nA + sB² / nB). Then t = (average of B − average of A) / standard error. This is Welch’s t-test, which doesn’t assume both groups have the same spread.

The degrees of freedom come from the Welch–Satterthwaite formula. With them, the two-sided p-value comes from the t distribution.

The 95% confidence interval is the difference ± critical t × standard error, with the critical t from that same distribution. The difference counts as real when the p-value is below 0.05. Each group needs at least 2 scores.

How to read the result

If the confidence interval doesn’t include 0, the difference is significant at 95%: it’s very unlikely to be due to chance alone.
If the interval includes 0, the difference could be chance. That doesn’t mean the groups are the same: you may need more responses to see it.
Look at the size of the difference too. With many responses, 0.05 points can come out significant and still not matter in practice.
If a group has fewer than 30 responses, take the result with care, especially if there are very extreme scores.

Frequently asked questions

Why do I have to paste every score and not just the average?: Because telling whether the difference is real requires knowing how much scores vary within each group. Two groups with the same average can have very similar or very scattered scores, and that changes the result.
What is Welch’s t-test?: It’s a way to compare two averages that works even when the groups have different sizes or different spread. That makes it a good fit for survey data.
What is the p-value?: It’s the probability of seeing a difference this large or larger if the two groups actually had the same average. If it’s below 0.05, the difference is called significant at 95%.
How do I get each person’s score?: With a satisfaction survey in each group. In SurvIQ you can create one for free from a template, share it by link, WhatsApp or QR code, and people respond without creating an account.

Know what you want to ask?

Build the survey in SurvIQ, share it by link, WhatsApp or QR code, and watch results come in live. The Free plan is free forever.

Create a free survey

More tools

NPS calculatorGet your Net Promoter Score from promoters, passives and detractors.eNPS calculatorFind out whether your team would recommend your company as a place to work.CSAT calculatorShare of satisfied customers and average score on a 1 to 5 scale.CES calculatorHow easy it was for your customers to get done what they needed.NPS comparison calculatorFind out whether the difference between two NPS scores is real or just chance.Weighted ranking calculator (branches or products)Rank branches or products without the one with 3 perfect scores taking first place.