Poll of the Day > What is with the discrimination in the PotD?

(+) Yeah! (+)

On June 24th, we had a poll about when a person normally wakes up. Today, June 25th, we have a poll about when a person normally goes to bed. The options for both only pertain to the day-time person. What about us night-shift people?
---

lightningbugx posted...

The options for both only pertain to the day-time person

But...both polls still gave you viable options. If you're a night shift person, then you likely go to bed at 6AM or later.

They're not going to list out every hour of the day for you. It's a junk poll.
---

@SBAllen why do you hate night shift workers

RCtheWSBC posted...

lightningbugx posted...

The options for both only pertain to the day-time person

But...both polls still gave you viable options. If you're a night shift person, then you likely go to bed at 6AM or later.

They're not going to list out every hour of the day for you. It's a junk poll.

They could list every 2 hours.
---

lightningbugx posted...

RCtheWSBC posted...

lightningbugx posted...

The options for both only pertain to the day-time person

But...both polls still gave you viable options. If you're a night shift person, then you likely go to bed at 6AM or later.

They're not going to list out every hour of the day for you. It's a junk poll.

They could list every 2 hours.

They actually can't, given that polls are capped at 10 options. They'd have to use a larger range than that if they wanted to cover the whole day. Rather than reducing the entire poll's resolution like that, it makes far more sense to adjust the bin sizes based on what are expected to be less common choices. That's common practice in making histograms, for the sake of making the data more readable. Nothing to do with discrimination.
---

In case you haven't figured it out, all the polls are just thinly veiled market research by CBSi.
---

Didn't time zones make the answers completely irrelevant? Also being a one day poll on a video game cheat site contributed to that.
---

Ah I remember the 12 hour night shifts I did. Good times and a lot more peaceful.
---

adjl posted...

They actually can't, given that polls are capped at 10 options. They'd have to use a larger range than that if they wanted to cover the whole day. Rather than reducing the entire poll's resolution like that, it makes far more sense to adjust the bin sizes based on what are expected to be less common choices. That's common practice in making histograms, for the sake of making the data more readable. Nothing to do with discrimination.

I see 11 options in those two polls. I think it actually caps at 12.

See, discrimination. Apparently it is okay to roll all of the night options into one category. Really, the data is not all that readable rolling a third of the day in one option and a third of the day into another option. You actually can't tell if the person is an early bird or night owl in those 2 categories. Thus, the data is still not readable.
---

lightningbugx posted...

Really, the data is not all that readable rolling a third of the day in one option and a third of the day into another option. You actually can't tell if the person is an early bird or night owl in those 2 categories. Thus, the data is still not readable.

This isn't a social science research center. GameFAQs' polls have rarely if ever been written to objectively collect data in a comprehensive manner. They are junk polls and aren't worth caring about.
---

RCtheWSBC posted...

This isn't a social science research center. GameFAQs' polls have rarely if ever been written to objectively collect data in a comprehensive manner. They are junk polls and aren't worth caring about.

So? They are fun polls, but sometimes these polls unnecessarily leave a segment of the daily visitors out.
---

lightningbugx posted...

Apparently it is okay to roll all of the night options into one category.

That is what I said, yes. When you expect a very tiny percentage of subjects to select a set of options, lumping those options together makes for more readable data sets.

lightningbugx posted...

Really, the data is not all that readable rolling a third of the day in one option and a third of the day into another option. You actually can't tell if the person is an early bird or night owl in those 2 categories. Thus, the data is still not readable.

It's perfectly readable. The first and last options could probably stand to be given upper and lower bounds, respectively (probably noon for both), but they still do a fine job of communicating "extremely early" and "extremely late," which is the point. 5.5% of voters go to bed after 6 AM, which is all it's looking for. That doesn't really need to be broken down into the <1% you'd get for each individual hour range between 6 and noon, especially at the cost of resolution in the range that encompasses 90% of voters.

lightningbugx posted...

but sometimes these polls unnecessarily leave a segment of the daily visitors out.

Sometimes, but this isn't one of those times. You are covered by the last option. Your vote just won't be discernible from that of somebody that goes to bed an hour earlier than you, which doesn't really matter. It also never matters even when the polls do actually leave people out, because the polls are completely meaningless.
---