Exam 10: Hypothesis Testing, Effect Size, and and Confidence Intervals: Two-Sample Designs

Here is the kind of problem Gosset was confronted with at the Guinness Company in Dublin. In the brewing process the yeast converts sugar into alcohol until the alcohol builds up to a level that kills the yeast. This typically happens at about 10-13% alcohol. The data below are for two strains of yeast. The numbers represent how resistant the yeast is to alcohol. Perform a t test at the .05 level, find d, and write a sentence summary of your conclusion. In addition, justify the statistical design you need to analyze these data. Strain A Strain B 10 14 11 13 12 9 9 13 9 13 11 14

For a two-sample t test, which of the following pairs of values qualify as large effect sizes?

Identify the following as paired samples or independent samples. In each case, give the appropriate df. -Eight volunteers each rated their emotionality in a large blue room and then in a small yellow one.

Identify for each description a) the independent variable, b) the dependent variable, and c) whether a paired or an independent-samples design is being described. -Linda conducted her independent project in Research Methods at a Post Office facility that sorted mail with zip codes that were not machine readable. She counted the number of pieces sorted on 8 days when a jam box was playing fast tempo music and on 8 days when it was playing slow tempo music.a) Independent variable _______________b) Dependent variable ______________c) Design ______________

Data Set 10-4: Gerbils were used to assess the effect of Natural Neutral, a drug designed to reduce emotionality in high-drive people. Each of 20 gerbils spent 10 solitary minutes in an open field. The investigator recorded the number of fecal boluses for each animal. Then each animal was given an injection of Natural Neutral and the open field task was repeated. The following data were produced (large numbers indicate high emotionality). No Drug Drug mean number of boluses 6 8 standard deviation of boluses 2 2 $\text { correlation coefficient between Drug and No Drug Score } = .50$ -The design in Data Set 10-4 is one of

This question requires careful thinking. The logic of null hypothesis statistical testing (NHST) involves assuming that

When statisticians talk about effect size, they are referring to

For each description, identify a) the independent variable, b) the dependent variable, and c) whether a paired or an independent-samples design is being described. -A 1957 article in the Journal of Experimental Psychology described the kind of dreams that are reported when you wake a person during REM sleep (rapid eye movement) or at other times during NREM sleep (non-REM). Vivid dreams come from the REM aroused and no dreams or vague dreams from the NREM aroused.a) Independent variable _______________b) Dependent variable ______________c) Design ______________

Data Set 10-6: Drugs used in treating schizophrenia all block the reception of dopamine by neurons. (Dopamine is a neurotransmitter, which, when released by the axons of one nerve, inhibits the firing of the next nerve.) This fact led to the idea that schizophrenia occurs when too much dopamine is produced. Suppose the following data on dopamine production were obtained. Schizophrenics Control Group 42 33 31 27 29 18 -The effect size index for Data Set 10-6 is

Here is an experiment modeled after that of Triplett (1897), who conducted the first social psychology experiment.Twenty-four 5th graders were given fishing reels. Twelve were tested individually. The number of yards of line reeled in 10 seconds was recorded for each. The other group was tested together under conditions of competition; the one who reeled the most line got to keep the reel. The following summary statistics were obtained. Individuals Group Competition \SigmaX 192 288 \Sigma 3347 7308 N 12 12 Analyze the results using a 99 percent confidence interval and d. Write a conclusion.

For each description, identify a) the independent variable, b) the dependent variable, and c) whether a paired or an independent-samples design is being described. -Pat's handwriting was terrible. Pat converted this longstanding problem into an independent project for his Research Methods class. The participants in his experiment graded an essay on a scale of 50-100. For 20 participants the essay was written in a terrible hand (Pat's own), and for another 20 the essay was written beautifully.a) Independent variable _______________b) Dependent variable ______________c) Design ______________d) df_______________

Data Set 10-5: This experiment is modeled after part of the data needed to demonstrate the Hawthorne Effect. Four assembly-line workers, who normally had two coffee breaks and an hour for lunch, participated in a week-long experiment in which they gave up both coffee breaks and half their lunch hour. Productivity (units produced per hour) for these four was measured under both conditions and is shown below. Productivity/hour Leisurely Schedule Rushed Schedule 6 6 3 5 4 5 5 6 -The experiment described in Data Set 10-5

You can increase the power of a statistical test by increasing

For a simple experiment, which of the following is true?

In one lab section of the Psychology of Learning course, students learned about operant conditioning by using a live rat. In a second section, students used a computer simulation. Afterward, both groups took the same 15-point quiz. Analyze the data using the appropriate techniques from Chapter 10 and write a conclusion about the two methods. Live Rat Computer Simulation 12 14 10 12 9 10 9 9 7 7 6 6 5

To be a small effect size, d must be .05 or less.

If a 95 percent confidence interval about a difference between means is 2.31 to 24.62, the null hypothesis may be

Identify the design of each experiment below as paired samples or independent samples. In each case give the df. -To control for socioeconomic status, this study used only families with annual incomes in the $30-40,000 a year range. Thirty-six fathers of one-year-olds were compared. Half had participated in the Lamaze method a year earlier when their child was born and half had not. A non-verbal behavior score was taken during a 10-minute father-child session.

Indicate whether each design below is independent samples or paired samples and where possible, give the degrees of freedom. -Alaskans with seasonal affective disorder (a kind of depression) spent 2 hours a day under one of two kinds of light-full spectrum or conventional fluorescent. At the end of treatment, depression scores were assessed for all those in the study.

According to your text, the reason we do experiments is to be able to tell