Re:Data Mining Hacker News: Front vs. Back
by statsknowitall on Tuesday October 11, 2016 @12:16PM
(#12685789)
Please don't use this language. Your suggested quote implies the null hypothesis has been confirmed. In fact, the null hypothesis has simply not been rejected. A better summary would be: "We did not find any proof that submitting articles at a certain time of day is an effective strategy to achieve front page visibility."
Re:Data Mining Hacker News: Front vs. Back
by WhitneyLand on Tuesday October 11, 2016 @12:35PM
(#12685964)
Yep, my mistake. Main point stands, helps to have some plain english.
Re:Data Mining Hacker News: Front vs. Back
by statsknowitall on Tuesday October 11, 2016 @12:59PM
(#12686206)
It's so easy to make mistakes when talking about this, with double and triple negatives that don't necessarily cancel out. I made one myself in the above comment before editing it.
Re:Data Mining Hacker News: Front vs. Back
by StavrosK on Tuesday October 11, 2016 @12:25PM
(#12685866)
Interesting observation, thank you. Very much changes the tone of the statement.
To that point, I read the full article and still don't fully understand the difference between fronts and backs. I have an idea of what it means after re-reading the description several times, but terms so central to the article should be made crystal-clear.
Absolutely agree. I think the most important part of being a data scientist is being able to communicate your conclusions (and their limitations!) to your audience. IMO, this is even more important than the actual analysis
One suggestion: For each important conclusion try to have at least one sentence that is understandable by a business exec.
For example at first glance it looks like time of day may be significant, then you conclude:
"Because the p-value is greater than the alpha value, we fail to reject the null hypothesis that the two nominal categories are independent."
By adding after this something like "Therefore submitting articles at a certain time of day is not an effective strategy to achieve front page visibility.", your post gains accessibility.
This is not a nitpick. The idea is to make sure the full power of your analysis is felt across a broad section of readers. Even if you send to tech people, these things often find their way to a wider audience.