Chapter 10 Two-Sample Designs t test Quiz 8

For a two-sample design, which of the following would be considered a small effect size? any negative number any negative number that is close to zero numbers that are close to 1.96 none of the descriptive alternatives are correct

any negative number that is close to zero

Which of the following has an influence on the power of a statistical test? sample size alpha actual difference in population means Correct! all of the descriptive alternatives are correct

all of the descriptive alternatives are correct

"blank depends on the number of observations minus the number of relations among the observations" is a statement about how to calculate Correct! degrees of freedom the difference between population means sx none of the other alternatives are correct

degrees of freedom - The formulae for the paired-samples t-test are on page 219 of your text, and the formulae for the independent-samples t-test are on page 220 of your text.

An effect size index that is considered small is: any negative value 0.20 1.00 1.96

0.20

The null hypothesis is: 1 = 2 1 2 1 > 2 1 < 2 any of the alternative answers are correct

1 = 2

In an independent samples design, the Dog's mean was 54.0 and the Cat's mean was 53.9. Larger scores are better. A t value of 2.50 was calculated and an a level of .05 adopted. Which conclusion is appropriate with a 2-tailed test? If df = 5, Dogs are significantly better than Cats. If df = 4, Cats are significantly better than Dogs. If df = 10, Dogs are significantly better than Cats. If df = 8, Cats are significantly better than Dogs.

If df = 10, Dogs are significantly better than Cats.

Which of the following statements is most closely related to confidence intervals? When extraneous variables are controlled, the experiment is valid. The difference between the two populations means was between .95 and .99. The difference between the two populations means was not zero. The mean of the first population is greater than the mean of the second population. The difference between the two populations means was between .95 and .99. The difference between the two populations means was not zero. The mean of the first population is greater than the mean of the second population.

The difference between the two populations means was between .95 and .99. - The formulae for the confidence interval for paired samples t-test are on page 210 of your text, and the formulae for the confidence interval for the independent-samples t-test are on page 213 of your text.

In which of the following would an increase from a smaller number to a larger number produce a more powerful statistical test? a β both of the descriptive alternatives are correct neither of the descriptive alternatives is correct

a

For a two-sample design, an effect size that is considered large is 0.80 1.00 1.50 all of the descriptive alternatives are correct none of the descriptive alternatives are correct

all of the descriptive alternatives are correct - The formulae for the effect size for paired samples t-test are on page 209 of your text, and the formulae for the effect size for the independent-samples t-test are on page 213 of your text.

The actual amount of difference between the two populations that the samples are from is most closely associated with: whether or not to use hypothesis testing Type I and Type II errors Correct! effect size effect size

effect size

The amount of difference that exists between two populations with unknown means (which you sample from) is the issue of: hypothesis testing or confidence intervals Type I and Type II errors the .05 or .01 a level effect size

effect size -

With an acknowledgment to Sesame Street, "Which of these things is not like the others, which of these things doesn't belong?" repeated measures natural pairs independent samples matched pairs

independent samples

As blank and blank increase, a statistical test becomes more powerful. actual difference between population means, the standard error of a difference sample variability, alpha sample size, beta none of the descriptive alternatives are correct

none of the descriptive alternatives are correct - Power is the likelihood that the null hypothesis will be rejected when it is in fact false. Power is affected by sample size, one versus two tails, standard error of the difference, and alpha level.

For a normally distributed set of scores, it is often best to use the design that has the most power. Which design has more power? paired sample independent sample both designs are equal in power Non-paired sample

paired sample

The null hypothesis in a paired samples t test is that the two: populations have identical means populations have different means populations have different means samples have identical means

populations have identical means - Paired samples designs involve two or more samples that have a dependent relationship between the samples such that the scores in each variable are systematically related to the scores in the other samples, so that these scores form pairs. Independent samples designs involve two or more samples that have an independent relationship between the samples such that the scores are not systematically related and do not have to be treated as pairs.

β is the blank of a statistical test. probability of a power both of the descriptive alternatives are correct neither of the descriptive alternatives are correct

power

If a 95 percent confidence interval about a difference between means is 2.31 to 24.62, the null hypothesis may be rejected at the .01 level rejected at the .05 level rejected at the .001 level retained

rejected at the .05 level

If a 95 percent confidence interval about a difference between means is -2.31 to 24.62, the null hypothesis may be: rejected at the .01 level rejected at the .05 level rejected, but more information is necessary before a particular level can be chosen retained

retained

Suppose an experimenter established a 95 percent confidence interval about the difference between two sample means and found values of -1.35 and +1.35. The null hypothesis should be rejected at the .05 level rejected at the .01 level rejected, but more information is necessary before a particular level can be chosen retained

retained

Sampling from populations that are normally distributed and have equal variances and using random assignment are considerations in deciding whether: to use a one- or two-tailed test you have enough power in your experiment the p value is accurate and your conclusion is correct to use hypothesis testing or confidence intervals

the p value is accurate and your conclusion is correct

If you have an independent-samples design, you can be sure that a Type I error is not possible a Type II error is not possible a confidence interval is not appropriate Correct! there is no logical reason to pair the scores in the two groups

there is no logical reason to pair the scores in the two groups

An experimenter found one sample mean of 13 based on an N of 8. The second sample mean was 18 based on an N of 6. The design was a paired-samples one Correct! was an independent-samples one could be either a paired- or an independent-samples one is not a two sample design

was an independent-samples one