Does CNN Know What Random Sample Means?

I titled this, “Does CNN Know What Random Sample Means?,” because when you conduct a poll your random sample is supposed to be representative of the population you are sampling, in this case U.S. registered voters. The poll they just did was conducted with 472 registered voters, 41% of them Democrats, 30% Republican, and 29% Independents. The poll showed that Mike Pence, a Republican, won the debate by 6% (48% vs. 42%), which CNN described as a, “narrow win.”

In a previous post, “Why the CNN Poll Cannot Be Trusted,” I described their poor polling strategy. Unfortunately they have not improved their ability to conduct an unbiased poll within the last week. Perhaps they could sign up for my $20 per month membership to figure it out. Either way they need to do something.

If you look at the voting demographics of U.S. registered voters you will see that there are 29% that identify themselves as Democrats, 26% that identify themselves as Republicans, and 42% that identify themselves as Independents at least according to this Gallup poll. If you were conducting a poll with a truly random sample you would expect to see similar numbers but unfortunately what we get is a much larger percent of Democrats than what is represented in the country. We also get a smaller Independent percentage and a somewhat similar Republican percentage.

This happened in their previous poll so I’m beginning to think these polls that CNN are conducting are not using a random sample. Shocker. If they do think they are, hey CNN, I’m available to help you.

O.k. so what can this poll tell us? Well 472 registered voters were sampled, roughly 227 thought Pence did better whereas 198 thought Kaine had the better night. The rest, I’m assuming (which is not a good thing to do), were indecisive as to who did better or worse.

What else do we know? Well out of those 472 registered voters roughly 194 (472 x 0.41 = 193.52) were identified as Democrats, roughly 142 (472 x 0.3 = 141.6) were identified as Republican, and roughly 137 (472 x 0.29 = 136.88) were identified as Independent. (I realize this adds to 473 so maybe there were 194 Democrats, 141 Republicans, and 137 Independents – it depends on how they rounded). These numbers really point out how underrepresented Independents were in this poll. They should be the largest group represented.

So what else can we make of this poll? Well Pence still won despite the Democrat bias shown in this poll. Outside of that do we really know what the 227 people are identified as? It’s probably a safe bet most Republicans thought Pence did better along with some Independents but at this point I’m making assumptions and that is not a good thing to do.

I’d really like to see a good random sample come from CNN. I hope they do better next time and hey, if you need help, contact me.

© Copyright 2016 I will solve that! All rights reserved. http://www.iwillsolvethat.com

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s