An analyst within your organization is presenting the result…
An analyst within your organization is presenting the results of a randomized experiment testing a new website design (treatment) against the current website (control) . They tracked the fraction of people who came to the website who “converted” (i.e., purchased, which takes values 0 and 1 with 1 indicating purchase). They calculated the treatment effect separately for three subgroups of users on your website. The graph below is the main output from the experiment. Please explain what you can learn and infer from this graph and what implications this might have for the decision about whether to launch the new website design.