Confidence Trend

[Mycal_Tamariki]

Hi all, 

We have a test that has been running for 2 weeks and it's at 92% confidence - I can see from the trend graph in A4T there are two points over the 2 week period where the highest points are above the current Confidence level which indicates to me that it has already peaked to 95% but we missed it.  I can't find a way in analytics to break down the confidence rate by day to see the actual percentages and it won't let me create a visualisation from the confidence dimension to achieve the same. 

There is no notification when a test reaches confidence and from what I can see it can decrease - should I leave the test to run longer or stop it claiming that at some point over the 2 week period it reached confidence even though we are at only 92% when it ended? 

Any help is appreciated

Many thanks

[alexbishop · Accepted solution]

I think the key thing is that the required sample size has been reached, did you calculate that before running the test? (For example using this calculator) It would then be a question of whether the required sample size had been reached by the time the 95% confidence level was reached