# Thread: Help with interpreting data regarding religious Affiliation and aspirations to cheat

1. ## Help with interpreting data regarding religious Affiliation and aspirations to cheat

An article is claiming that likelihood to cheat can be predicted by religious affiliation. I have provided some of their data on their sample of aspiring cheaters. I gathered data on religious affiliation in the United States. Given the following information, what can we conclude.

Religious affiliation in the United States(Rounded)(I realize this doesn't make up 100%)

Evangelical Protestant 26%
Catholic 24%
Mainline Protestant 18%
Muslim .6%

Religious Affiliation of Aspiring cheaters(Data taken from a website comprised of over 60,000 members of people aspiring to cheat on their spouses)

Evangelical Protestant 25%
Catholic 23%
Mainline Protestant 23%
Muslim 1.5%

I will put them in in pairs for convenience.(Religious Affiliation in U.S/Religious Affiliation of aspiring cheaters in U.S)

Evangelical Protestant 26%/25%
Catholic 24%/23%
Mainline Protestant 18%/23%
Muslim .6%/1.5%

My novice level of statistics says that we can make some general predictions based on the above data. For instance, if i was to walk up to a stranger I could say that they are most likely to be an Evangelical Protestant and least likely to be Muslim. As for the data on aspiring cheaters, since the sample size is so large, we can (somewhat) assume that this data represents all aspiring cheaters in the United States. For instance, if i walk into a room of aspiring cheaters and approach someone randomly I could assume that they are more likely to be Catholic than Mainline Protestant.

Assuming I am correct with the above assumptions, what other conclusions can be made? What about the cases where the the percentage of religious affiliations of aspiring cheaters is higher than the percentage of the same affiliation in the US, as we see with Mainline Protestants and Muslims? I am not sure what the disparity between the sample of aspiring cheaters and the U.S. population means. Can we make any assumptions of someone's likelihood to cheat based on religious affiliation using the above data?

2. ## Re: Help with interpreting data regarding religious Affiliation and aspirations to ch

Hi,
you could test the null hypothesis, that there is no link between religion and aspiration to cheat vs. the alternativ hypothesis that there is a such a link. Yu would need to make sure your percentages add up to 100% though. My guess is that you would need to introduce and measure a further category - something like "unknown releigion" or "Other" - it is a quite a large chunk of yur population.

Once you have this you might run a chi-square goodness of fit test to check the link between religion and cheating .

A further warning - establishing a link does emphatically NOT mean that certain religions predispose people more towards cheating - the link might have a lot of different explanations. E.g. the muslims in your sample will probably stand out - but there can be many biasing effects to explain your finding apart from religion.

regards
rogojel

3. ## Re: Help with interpreting data regarding religious Affiliation and aspirations to ch

(Data taken from a website comprised of over 60,000 members of people aspiring to cheat on their spouses)
since the sample size is so large, we can (somewhat) assume that this data represents all aspiring cheaters in the United States.
Only those who use such a website.
I am not sure what the disparity between the sample of aspiring cheaters and the U.S. population means.
Those mainline protestants who want to cheat are more
likely to use such a web site than other religious
affiliations?

With kind regards

K.

4. ## The Following User Says Thank You to Karabiner For This Useful Post:

thecosmodreamer (06-07-2014)

5. ## Re: Help with interpreting data regarding religious Affiliation and aspirations to ch

You are totally correct. I made a mistake assuming i could make assumptions about all cheaters based on this sample of the population. This data would need to be taken from a random sample of all people in order to make those generalizations. Still, we could only make assumptions about an aspiring cheater's religious affiliation, not the other way around.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts