Interested in the correlation
Fritz O'Doul
 9/22/2016 1:56:00 PM      Replies: 32

Fritz O'Doul9/19/2016 11:14:29 AM

Hi All!

Exciting (scary?) times here for GW Basketball! I want to test out a theory here regarding the correlation between the opinions on this board regarding the Lonergan firing, and political leanings. Please fill out the below survey, and I will release results after they come in:

https://goo.gl/forms/i9URi64MhPCOrsna2

Best,

Fritz O'Doul

 

24.102.114.2319/19/2016 1:47:02 PM

Done.

NYCOLONIAL9/19/2016 1:50:20 PM

Done. Interesting idea.

Craig Martinez9/19/2016 2:46:59 PM

Done

24.102.114.2319/19/2016 3:35:02 PM

Done and Bump

greenpoint ny9/19/2016 3:38:00 PM

done

Bigfan9/19/2016 3:40:51 PM

People think this is a good idea?

Fritz O'Doul9/19/2016 4:48:56 PM

I am getting some interesting data! Keep the responses coming and I will share insights! 

107.77.76.1119/19/2016 4:56:04 PM

Any correlation is likely bogus because the sample is likely of a largely left leaning population. To make this valid you would have to have a fair representation of Dems Reps and Independents. There is no correlation ... garbage in equals garbage out.

Dootie Bubble9/19/2016 5:02:23 PM

As someone who does surveys as part of his job the issue is representation of your sample not really size.  As long as the Dems and Reps are reasonably like the average Dem and Rep GW fan results should be good.

Long Suffering Fan9/19/2016 5:47:29 PM

Although I prefer to separate GW basketball and politics. This is an interesting survey and am curious to see results.   My one minor gripe is until we see if anyone is leaving, it is hard to get a read on how the win total may be impacted

NYCOLONIAL9/19/2016 5:49:39 PM

Fritz never claimed this was a scientific sampling.  Is there anybody around here who doesn't think we could all benefit from a little fun right about now?

The Sons of Liberty9/19/2016 9:58:39 PM

I wasn't going to participate until I saw that others were doing it.

You can add that to the calculations.

 

Rich Maier9/19/2016 10:04:17 PM

Done

BC9/19/2016 10:36:30 PM

why do allow more than one submission, seems unreasonable to me.

GW699/20/2016 8:40:03 AM

I was terrible at statistics so help me out.If you can submit as many 

times as you like--like BC points out---and can submit without a handle,

or under numerous handles or your handle as often as you wish,

doesn't that skew the results?I needed to obtain a book called 

"Intuitive Statistics"and I still didn't get it.

CJS Fan9/20/2016 11:22:50 AM

Done

Dootie Bubble9/20/2016 11:23:04 AM

You bias your results if your sample is not representative of your population. Allowing anonymous trolls on the Internet to post multiple entries is not industry standard for collecting a reasonably representative sample. In other words I wouldn't alter my vote or position on ML based on the findings.

ziik9/20/2016 11:42:42 AM

This population? How can it be represented? Nerdly hangers on?

GW699/20/2016 11:52:04 AM

Thanks Dootie and Ziik.

BC9/20/2016 12:05:36 PM

Any internet poll has got to be viewed with suspision.   It's a biased sample to begin with.

BC9/20/2016 12:15:03 PM

The only problem with quotes (and polls) on the internet is that you can't authenticate them -- Abraham Lincoln

Bigfan9/20/2016 12:21:51 PM

Think it's odd and simplistic, but "Fritz O'Doul," whomever he is, portrays it as fun. And we can use fun.

Spoiler alert: if answered honestly and non-repetitively, poll will find Trump supporters are mostly against firing, Clinton supporters mixed. If Bernie Sanders were in the equation, most of his supporters would like support the firing.

Doesn't take a political expert to figure it out.

BC9/20/2016 12:37:04 PM

You make too many assumptions.

The Dude9/20/2016 2:34:26 PM

Fun idea Fritz.  The Q you pose had crossed my own mind.  

73.196.210.2289/22/2016 8:44:41 AM

I am guessing the results will be posted as soon as 

GW names a new coach.  Procrastination must be in the water.

Fritz O'Doul9/22/2016 10:39:41 AM

Haha sorry yall I was on a business trip the last two days, which is good, as the survey has racked up about 90 responses. I recognized folks may have double submitted to skew the results, but that seems fairly unlikely to me. I will post results and some observations in the later afternoon today!

JPW9/22/2016 10:46:55 AM

When do you expect to post some results?

NYCOLONIAL9/22/2016 10:52:24 AM

Fritz, it is interesting that you find it fairly unlikely that some people would double post. Personally, I think some of us believe that those who favor the OTHER candidate are far more likely to cheat. And I do admit to my suspicions, particularly since these boards are also open to the non-GW community and there's no telling who might "vote," and how many times.

NYCOLONIAL9/22/2016 10:53:11 AM

Having said that, I still think it's a fun exercise, and we could certainly use some fun, cuz there hasn't been enough of it lately.

Fritz O'Doul9/22/2016 1:25:35 PM

Some Basic Stats and observations:

Total Respondents – 95

A note on duplication: It is challenging to account for duplicates, even though I can see timestamps. It also seems unlikely that one side of any question here would have more people submitting multiple entries as the other. There were only two instances of exact duplicate entries being made within a small (5 minute) window. That said, it is a statistical issue and makes these results little more than fun and interesting.

"Do I believe that Mike Lonergan should have been fired"

Based on the responses, it is not very surprising that tensions have been fairly heightened in lieu of recent events:

Yes: 37.9% (36)

No: 36.8% (35)

Undecided: 25.3% (24)

 

“Who am I supporting for POTUS:

Hillary Clinton: 67.4% (64)

Donald Trump: 15.8% (15)

Other: 16.8% (16)

Unsurprising given that GW is the thread between all respondents and the voting tendencies of Washington DC, that Hillary is doing substantially better than Trump. One can safely assume that many of the respondents are College Graduates a population among which Clinton soundly beats The Donald. The most interesting point here may be the 16.8% of respondents supporting another candidate, (likely Jill Stein or Gary Johnson).

 

“Estimate how many wins you believe will turn to losses due to the firing of the head coach”

Mean estimated losses added: 4.95

0: 6.6% (6)

1: 1.1% (1)

2: 9.9% (9)

3: 13.2% (12)

4: 11% (10)

5: 24.2% (22)

6: 12.1% (11)

7: 4.4% (4)

8: 6.6% (6)

9: 1.1% (1)

10: 8.8% (8)

14: 1.1% (1)

Did not respond (4)

Interesting note: 55 respondents or 60.44% believe that the impact will fall between 3 and 6 wins converted to losses, with 5 being a pretty clear concensus.

 

Deep dive and correlations:

Of Trump supporters:

20% Think Lonergan shouldn’t have been fired (below the average for the wider sample)

33% Are undecided (above the average)

46.67% Think he should have been fired (above the average)

I find this to be a surprising result (though I recognize the limited sample size), as I had hypothesized that Trump’s sort of aversion to ‘political correctness’ would have meant more supporters felt that the coach should have been retained. I would be interested in what a larger sample might show.

Of Clinton supporters:

37.5% Think Lonergan shouldn’t have been fired (right about the average for the wider sample)

21.88% Are undecided (below average)

40.63% Think he should have been fired (slightly above the average)

Of other voters:

50% Think Lonergan shouldn’t have been fired (well above the average for the wider sample)

31.25% Are undecided

18.75% Think he should have been fired (Well below average)

The correlations here with non Trump/Clinton supporters are quite interesting, and im not sure what to make of them.

I can do a bit more analysis later and try and post datasets (with any handles ommitted). Let me know if you have specific questions!

JPW9/22/2016 1:55:59 PM

thanks

      Stuff you should read

  • Make an argument
  • Don't call someone an evil pant-load
  • Don't threaten to sue someone for your free, voluntary participation on a semi-anonymous site

 Respond

Thread Stats

Active Responders

Timeline

  • Most active day: 9/19/2016 (15 / 45.45%)