This preview shows pages 1–4. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: What sample size should we use? Rule of thumb: When comparing two groups, keep the SE for each group substantially smaller than onefourth of the anticipated difference between the two group means. Example: If you are comparing soybeans grown under two environmental conditions, and you expect stem lengths to differ by about 0.8 cm, then you would design the experiment to keep SE less than 0.2 cm (as above). desired SE = guessed SD n n = 13 y = 21.34 cm s = 1.22 cm SE = .34 cm Suppose we want to reduce SE to .2 cm. We’ll use s =1.22 cm (from the old study) to guess the SD SE = 1.22 n ≤ .2 .2 n ≥ 1.22 n ≥ 1.22/.2 n ≥ 37.2 use at least 38 plants Soybeans : When can we use the Student’s t method to construct confidence intervals? 1. Conditions of the study design: (a) It must be reasonable to regard the data as a random sample from a large population. (The sample should not be more than 5% of the population.) (b) The observations in the sample must be independent of each other. 2. Conditions of the form of the population distribution: (a) If n (sample size) is small , the population distribution must be approximately normal . (b) If n is large , then the population distribution need not be approximately normal . Why? (a) It must be reasonable to regard the data as a random sample from a large population. Example: 10 shortsighted undergrad students volunteered for a psychological study, and were found to have mean IQ of 128.4, whereas mean IQ for the general population is known to be 100. Can we conclude that this is a general phenomenon? Let’s look closer: The 10 individuals in the study are undergrad students. Is this a random sample from the population in general? No! Since their decision to join the group might be related to their intelligence, you probably cannot regard them as a random sample (with respect to IQ)....
View
Full
Document
This note was uploaded on 11/04/2009 for the course BIO 50935 taught by Professor Bryant during the Fall '09 term at University of Texas.
 Fall '09
 BRYANT

Click to edit the document details