layout | title | nav_order |
---|---|---|
default |
Run the test |
3 |
{: .no_toc }
A simple web app that helps identify interactions between tests. {: .fs-6 .fw-300 }
To test for interactions between two experiments 1
and 2
, we will need
counts for each combination of variations of the two tests. The first field
below labeled 1A+2A
should contain the count for visitors who were exposed
to the first variation (A
) of experiment 1
and also to the first variation
(A
) of experiment 2
. Visitors who were exposed to only one of the two tests
are not included in this analysis.
{% raw %}
Conversions in 1{{ String.fromCharCode(65 + index) }}+2{{ String.fromCharCode(65 + cindex) }}
Warning. Possible interaction detected with {{ display_p }}.
The observed sample sizes in each group does not match the expected proportion of the total sample size.
Comparative statistics may be invalid as a result.
<div class="alert alert-success text-center" v-if="!flag_interact">
No indication of interactions with <var>{{ display_p }}</var>.
</div>
</div>
</div>