# PSYC 3000 Lecture Notes - Lecture 6: Sampling Distribution, Chi-Squared Distribution, Level Of Measurement

18 views2 pages

2 Aug 2016

School

Department

Course

Professor

PSYC 3000

Sept 24

Hypothesis Testing

-hypothesis are always made about the population

-use sample to test it.

-sampling distribution is the probability of different outcomes for this experiment

Test of the mean

-step 1: statistic chosen=mean

-make a hypothesis about nature of the mean in population

-measure the mean in the sample.

-every time you take a sample, you look at the sample of the mean. Eventually you will look at sampling

distribution of the mean.

-leptokurtic=thin and pointy

-platykurtic=flat distribution

-the way to create a sampling distribution is to take many samples from the same population. The mean

won’t be the same for every sample.

-after many samples have been taken the shape of the distribution will not move around much.

Test of the skew

-hypothesis a skew in a population (ex: 0)

-take many random sample and calculate the skew from each.

-end result is always conclusion about the population.

Test of the variance

-called the chi square test

-will always give a chi square distribution.

-because the variance can never be 0 your hypothesis can never a normal distribution but always a

positively skewed one.

Test of a proportion

-when testing nominal data you change it to proportions. You would estimate that a proportion of the

population

-when asking a “yes/no” question you must give yes and no a numeric value

Test of the mean

-a test of pairs of means could be carried out to find out if the difference is significant

-every time a test is listed in a paper it will have a p value.

P values

-different p values will yield different data

-whatever your alpha level is it exists independent of your data. Your alpha level is chosen before your

data is collected.

-p values arrive out of your data.

find more resources at oneclass.com

find more resources at oneclass.com