1. A hypothesis is a guess about a population statistic, such a mean.
a. If technological improvement increases productivity, then the mean
of growth will be greater than 0.
2. We start by asking what a null hypothesis would look like.
3. We want to be 95
1. Probability sample: every response ahs an equal chance of being randomly selected. a. Keep doing random sample we get a different number but similar. 2. Sample mean is a random variable. This signifies that the sample means is not fixed but varies. Var
1. Standard Normal Distribution: Is symmetric has half on each side of mean.
a. We will compare our data distribution to the Normal Distribution.
2. Skewed distribution will have a lot of observations further to one side than
a. Right skewed mo
1. Describing varations
a. Interquartile range: is the distance between the 25 th to 75th percentile.
The smaller distance means more people fall around the mean/median. 2. Box Plot
a. Shows the 25th, 50th, and 75th percentile. b. Shows outliers=1.5*IQR f
1. Proportion of people who agree is people who agree over the total number
of respondents. The percent is the proportion multiplied by 100.
2. Are there systematic differences between two groups
a. This is why we tabulate cross variable comparisons
a. Stratified random sample
i. Break into groups do probability sampling within groups.
ii. At the end proportion of each group matches population
proportion. Overrepresent in the sample that control in final
data, but have enough data to also make conclu
1. Good sample have characteristics a. Representative of the underlying population. Looks like population. Don't only want to look at males or females and apply it to everyone.
b. Large enough to draw reliable inferences. Can't have two people represent U
a. Running records: i. have lower costs ii. More superficial iii. May be systematically altered iv. Increased sample size v. Government data can sometimes be of poor quality.
2. Challenge of Sampling: want to be sure the measurements from a sample reflect
1. Factors in deciding type of observation
a. Cost: cheaper is better. Some techniques are cheaper than others.
b. Availability: are you going to be able to gain access to the
individuals you want to talk to.
c. Underreporting: some events tend to be unde
1. Primary v. Secondary
a. Primary data is data that the researcher themselves create.
i. Creating a transcript from an interview. Results from an
experiment you created.
ii. Primary because you are the one doing it and directly
observing what is going on
1. Segal and Cover: measure ideology of SCOTUS justices a. Looked through newspaper editorials written about each justice. Based on these editorials, can figure out how liberal or conservatives. b. Looking at editorials in major papers, rate each paragrap