Search in the document preview
HPER-h 391 Statistics Epidemiologic Testing & Independent t-test
Epidemiology: reasoning and inferences about disease, examines the distributions and causes.
Rate: Number of Health events/population at risk
Measured by incidence (new illness) and prevalence (existing illness)
Odds ratio: (a/c)/(b/d)
Relative risk (a/(a+b))/(c/(c+d))
Previous tests compare the sampling mean with the population mean, but many times we do not know the population mean. Instead it is best to use tests between means of two groups that meet the following criteria:
• Not related in any way
• Each participant is tested only once
• Assumed: homogeneity of variance (assumed to be equal)
• Effect of violating this assumption is not serious if n1=n2 or a large sample size is used
• Obtained value= statistic-parameter/standard error
• Standard error is found using the square root of s^2pooled.
S pooled value= everything under square root sign.
SPSS Instructions: Make a group variable using values 1 and 2 then enter data
Analyzecompare meansindependent sampling test
Test variabledependent variable
Look at: Levine’s test for variance (should be p>.05) This ensures validity!
Sig: if >.05 accept Null Hypothesis. If it is <.05 reject the null hypothesis.
Also make sure you look at the obtained value. Always check your calculations with the SPSS outputs!
Dr. Love wants see if not eating breakfast helps competitive eaters eat more hot dogs. He compared the data from two separate contests. In contest 1 the eaters had eaten breakfast before the contest. In contest 2, the contestants had not eaten breakfast.
Alpha level: .05
Number of hotdogs eaten:
Contest 1: 12, 13, 5, 7, 2, 12, 14, 13, 22, 4, 3, 21, 7
Contest 2: 6, 8, 22, 19, 4, 21, 28, 26, 22, 3, 13, 14, 15
Now use SPSS to calculate the t value and compare answers!