SOC202H1S Qualitative Analysis in Social Sciences
January 30 2012 – Week 4
Today’s lecture will be a bit more abstract and a lot of materials will be from the
- There is 1% of population that is very extreme. Most people are centered
because that is what average is about. For example, some people are really
rich but not all. Essentially what we will do is we will take that idea – take
sample statistics and see if it is representing a population by various tests.
Sleep: in a typical 24-hour period, approx. how many hours of sleep do you
- There are very few people that have depression. If depressed people were
common, our result will be very different.
- In the test he will get you to think about distribution and examples. So he will
have an example of huge Canadian survey. So what would happen to the
mean if we added 1 more case to 1 hour? Then the data wont change so
much, because sample size is so huge
- Larger the sample, it absorbs amount of potential of outlier
- The more deviant from the norm you have, it is very easy and political to
- One case of deviance is not a big deal. But if there is a lot, it becomes the
o You are trying to get average income, but person that makes 50,000
dollars a day can pull it up.
- Standard deviation – seeing how far each score is from the mean.
- The relationship between hours of sleep and conflict
- We will establish the relationship of X and Y here.
- People are not truly representative. People who are here and not here hold
different chance to be selected in the sample.
- How likely is it that we can grab 2 samples and obtain different result?
- Say we scoop another sample on top of a sample, we might be get different
result. But if it was done correctly and appropriately, the both results should
be fairly close to each other. If they are significantly different, there is
- Figure 7-1
o One sample is off by 4.0, 5.5, 4.3, etc. it is the estimate of the idea
o This is the idea behind the notion of sampling
- We basically want to quantify an error. Use as template as a sample of
- The bigger samples are better because there