Q&A Hero

Q: 53 A researcher hypothesized that adolescent girls are more invested in their romantic relationships than are adolescent boys. She asked adolescent partners to rate the total amount of time they engaged in activities centered around their relationship (e.g., seeing each other, thinking about the partner, talking about the relationship with others). In a sample of 41 couples, girls spent an average of 12 hours a week and boys an average of 10. The standard deviation of the difference was 1.5 hours. a) Calculate t. b) Was the research hypothesis supported? Explain.

Q: 52 A kindergarten teacher wanted to see if her students' social skills changed over the academic year. She compared the scores on a standardized measure of social skills at the beginning of the year, and the end of the year. The results follow: a) What is the null hypothesis? b) Is a one or two-tailed test more appropriate? c) Calculate t. d) Write a sentence explaining the results.

Q: 51 Give an example in which carryover effects may impact a repeated measures design unduly.

Q: 50 Generate three examples of research questions that are appropriate for testing with a related samples t-test.

Q: 49 In repeated measures designs, the difference score refers to the difference in standard deviations for the two variables being compared.

Q: 48 A repeated measures t test would be appropriate to see if students who took the SAT in 2000 scored higher than students who took the SAT in 2002.

Q: 47 The formula for calculating a related samples t includes the standard deviations of the two variables being compared.

Q: 46 In a related samples t test, the difference score is usually compared to 0.

Q: 45 The variables under study in repeated measures designs are unrelated to one another.

Q: 44 Individual differences between subjects are less problematic in related samples designs than in independent samples designs.

Q: 43 Related samples designs require fewer subjects than independent sample designs to achieve the same degree of statistical power.

Q: 42 Repeated measures designs are subject to order effects.

Q: 41 The degrees of freedom in a related samples t-test are N - 2.

Q: 40 The scores of the same students on a pre-test and a post-test are independent.

Q: 39+ Cohen's d refers to a) the probability associated with t. b) the difference between the means before and after treatment. c) the "diagnostic" statistic. d) the difference between the means before and after treatment divided by a standard deviation.

Q: 38+ The example of the moon illusion discussed in the text illustrates the fact that a) the best estimate of the size of an effect need not use the standard deviation. b) standardized effect sizes are clearly superior. c) the moon grows as it rises in the sky. d) an effect size is best based on medians.

Q: 37 When we have related samples, the best measure of the effect size uses a) the standard deviation of the difference scores. b) the standard deviation of the pretest scores (if they exist). c) the pooled estimate of the pre- and post-score standard deviations. d) none of the above

Q: 36 Suppose that we take 15 gay couples and observe the difference within couplesin terms of age. Then we take 15 straight couples are record the same differences. We want to test if straight couples are more similar in age than gay couples. (There is some reason to expect that this is true.) What statistical procedure would be most appropriate? a) computing a correlation coefficient for each couple b) running an independent samples t test between gays and straights c) running a t test for related samples d) using a chi-square test

Q: 35 If the experimenter had instead used an independent samples design with the same number of participants a) the power of the design would have increased. b) the power of the design would have decreased. c) the power of the design would have stayed the same. d) the power of the design would not be predictably affected.

Q: 34 An experimenter collected data on how well a study guide improved grades on an exam taken late in the semester compared to an exam taken early in the semester. Using a related sample means t test, the results showed that later grades were higher than early grades (t(74) = 3.64, p < .05). Which of the following was NOT an advantage of this design? a) The design controlled for students who did poorly both times. b) The design controlled for extraneous variables like intelligence levels. c) The design controlled for carry-over effects from already having taken one exam when the second exam was administered. d) All of the above are advantages of this design.

Q: 33+ The null hypothesis of a related scores t test is a) b) c) d)

Q: 32 The difference between the values of degrees of freedom for one samplet tests and related means t tests is thata) related means t tests have a df = N - 2, where N is the number of pairs of scores.b) related means t tests have a df = N, where N is the number of pairs of scores.c) related means t tests have a df = N - 1, where N is the number of pairs of scores.d) one sample t tests have a df = N - 2, where N is the total number of raw scores.

Q: 31+ As the value of the mean difference score decreases a) thet score increases. b) thet score decreases. c) thet score stays the same. d) You cannot predict how the t score will be affected.

Q: 30 If two sets of measures have the same mean, but different variances, the resulting t will be closest to a) 1.00 b) 3.00 c) 0.00 d) It is impossible to know.

Q: 29 The standard error of the difference between two means is a) the standard deviation of a set of difference scores. b) the standard deviation of a set of means of difference scores. c) the variance of the means. d) the standard deviation of the pretest scores.

Q: 28 In the t test for repeated measures the symbol stands for the a) standard error of the mean. b) standard error of differences between means. c) standard deviation of differences scores. d) There is no way to know.

Q: 27 The t test for two related measuresa) is complicated by the fact that we have two different sets of numbers.b) is simplified by the fact that we really only focus on the column of difference scores.c) is not a valid test if the pre- and post-measures are correlated.d) is impractical because you do not know the population variance.

Q: 26 In the Kaufman and Rock (1972) moon illusion example in the text, they hypothesized that there would be no moon illusion in their experiment. Experiments of this type pose problems for researchers because a) you cannot logically prove the null hypothesis to be true. b) the null hypothesis is obviously true. c) rejecting the null hypothesis would mean that Kaufman and Rock were wrong. d) the null hypothesis is obviously false.

Q: Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperELEVATED - LEVEL.0190.13714.3E-02-.0791.1171.4389.67225 If the effect of the first measurement influences what the subject does on the second measurement, we would name thisa) a treatment effect.b) a carryover effect.c) a contaminating influence.d) a flaw in the design.

Q: Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperELEVATED - LEVEL.0190.13714.3E-02-.0791.1171.4389.67224+ The standard error of the mean would normally be writtena) 4302b) 0.043c) 430d) 43

Q: Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperELEVATED - LEVEL.0190.13714.3E-02-.0791.1171.4389.67223 The confidence limits in the output can best be interpreted to meana) the difference between the pre- and post-test sample means is between -.0791 and .1171.b) the population mean must be at least =.0791.c) an interval computed in this way has a probability of .95 of encompassing the difference in population means.d) the population mean has a probability of .95 of lying between -.0791 and .1171.

Q: Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperELEVATED - LEVEL.0190.13714.3E-02-.0791.1171.4389.67222+ The 2-tailed significance level tells usa) the difference is not significant at the .05 level.b) the difference is significant at the .05 level.c) the means of the pre- and post-scores are large.d) We can"t tell whether the difference is significant or not.

Q: Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperELEVATED - LEVEL.0190.13714.3E-02-.0791.1171.4389.67221 In the output, the value of .1371 stands fora) the standard deviation of the pre-test scores.b) the standard deviation of the post-test scores.c) the standard deviation of the difference scores.d) none of the above

Q: 20 The standard error of the mean of difference scores could be calculated bya) taking the standard deviation of the differences.b) repeating the study many times and looking at the distribution of means.c) dividing the standard deviation of difference scores by the square root of the sample size.d) looking at the standard deviation of the posttest scores.

Q: 19+ Which of the following is sometimes a serious problem with repeated measures designs? a) Carryover effects can cloud the interpretation. b) Small sample sizes can distort the results more than with other designs. c) They require more subjects than designs with independent samples. d) all of the above

Q: 18+ Which of the following was NOT an advantage of repeated measures designs discussed in the text? a) It allows us to avoid problems associated with variability from subject to subject. b) It helps to control for extraneous variables. c) It is easier to calculate the statistics. d) It requires fewer subjects that other designs.

Q: 17 Which of the following are reasons why we might NOT use a repeated measures t? a) It requires too many subjects. b) It is more likely to reject a null hypothesis than the design with difference subjects in the groups. c) Information the subjects pick up in early trials may influence their performance on later trials in ways that we don"t find helpful. d) It allows the correlation between trials to influence the results.

Q: 16 In a repeated measures t, the degrees of freedom are equal to a) N. b) N " 1. c) N " 2. d) the number of observations in the two conditions.

Q: 15+ If the critical value of t associated with the above formula is 2.12, what would you conclude about your means? a) There is a significant difference between the means. b) There is a significant difference between the standard deviations. c) There is no significant difference between the means. d) p > .05

Q: 14+ In the formula for t, 1.73 is a) the standard deviation of the sample. b) the standard deviation of the population. c) the standard error of the mean. d) the difference between the means.

Q: 13+ In the formula for t, there are _______ pairs of observations in the study. a) 17 b) 34 c) 18 d) 4.18

Q: 12+ At test, in general, involves a) dividing the difference between means by the standard deviation of the population. b) dividing the difference between means by the standard deviation of the sample. c) dividing the difference between means by the standard error of a distribution of differences between means. d) dividing the difference in standard deviations by the size of the larger mean.

Q: 11+ The null hypothesis in a repeated measures t test is a) the hypothesis that the mean difference score is equal to 0. b) the hypothesis that the mean difference score is different from 0 in either direction. c) the hypothesis that post scores are larger than pre-scores. d) the hypothesis that the variance of scores stays constant from pretest to posttest.

Q: 10+ A repeated measures t test is more likely to lead to rejection of the null hypothesis if a) subjects show considerable variability in their change scores. b) many subjects show no change. c) some subjects change a lot more than others. d) the degree of change is consistent across subjects.

Q: 9 If we test the mean amount that alcoholic subjects drink before and after therapy, and that difference is NOT statistically significant, this could mean a) the therapy was not effective. b) the sample size was too large. c) the study lacked sufficient power. d) a and c

Q: 8 We are evaluating a method of therapy for extremely underweight adolescent girls. If we weighed our subjects at the beginning and end of therapy, a difference in weight could mean a) that our therapy worked. b) that people gain weight over time regardless of what we do. c) that our scales changed due to repeated use. d) all of the above

Q: 7 The mean of a column of difference scores is equal to a) the ratio of the means of the individual columns. b) the difference between the means of the individual columns. c) the sample size. d) We can"t tell without calculating it for a set of data.

Q: 6 A difference score is obtained by a) subtracting the Before score from the After score. b) subtracting the After score from the Before score. c) dividing After scores by Before scores. d) either a or b, just so long as you are consistent

Q: 5 We would be least likely to use a repeated measures design when a) there are substantial individual differences. b) there are minimal individual differences. c) we want to control for differences among subjects. d) we want to compare husbands and wives on their levels of marriage satisfaction.

Q: 4 In the preceding question on autonomy in children, we would be most likely to use that design, rather than random sampling of children, because a) we want to control for differences in means. b) we want to control for differences in parenting style. c) we expect scores of children in the same family to be unrelated. d) we want to control for differences in age between first and second born children.

Q: 3+ We want to study the mean difference in autonomy between first-born and second-born children. Instead of taking a random sample of children we take a random sample of families and sort the children into first- and second-born. The dependent variable is a measure of autonomy. This experiment would most likely employ a) a repeated measures analysis. b) an independent measures analysis. c) a correlation coefficient. d) a scatterplot.

Q: 2 We treat the repeated sample case differently from the case involving two separate samples because of a) the difference in the means of the two samples. b) the fact that different subjects were involved. c) the correlation between the two sets of data. d) the size of the sample.

Q: 1 Which of the following terms does NOT belong with the rest? a) related samples b) repeated samples c) independent samples d) matched samples

Q: 68 Briefly describe two factors that affect the magnitude of t.

Q: 67 Explain the following statement: p (100­110) = .90.

Q: 66 In the previous question, what would be the minimum mean score of the teacher's students that would yield a statistically significant difference using a one-tailed test?

Q: 65 The average SAT score for a local high school was 1100. One teacher is convinced that the 25 students who were in his homeroom performed better than the average student in the high school. Their average score was 1125 with a standard deviation of 100. a) Calculate t. b) Evaluate the teacher's "hypothesis" in light of t.

Q: 64 The mean anxiety score in elementary school children is 14.55. A researcher wants to know if children of anxious parents are more anxious than the average child. Below are the anxiety scores from 10 children of anxious parents. 13 14 14 15 15 15 16 17 17 18 a) Calculate the t value. b) Write a sentence to answer the researcher's question.

Q: 63 The following 10 numbers were drawn from a population. 5 7 7 10 10 10 11 12 12 13 a) Calculate the 95% confidence interval for the population mean. b) Is it likely that these numbers came from a population with a mean of 13? Explain.

Q: 62 Calculate the 95% confidence interval for ­ given s = 25, and N = 101.

Q: 61 Given ­ = 100, s = 27, and N = 30:a) Calculate t.b) Write a sentence interpreting the value of t as a two-tailed test.c) Write a sentence interpreting the value of t as a one-tailed test.

Q: 60 Given a sample size of 30, and one sample t = -2.5, what would you conclude about the sample from which the mean was drawn?

Q: 59 Assuming a two-tailed one sample test is being used, what are the critical values for t given the following sample sizes: a) N = 10 b) N = 15 c) N = 30

Q: 58 The larger the difference between the sample mean and the population mean, the larger the t value.

Q: 57 As the sample variability increases, the magnitude of t increases.

Q: 56 When comparing the mean of a sample of 30 people to the population mean, the degrees of freedom are 31.

Q: 55 Student'st distribution essentially accounts for the fact that t is often larger than the corresponding z because it is based on estimated variance, which is biased.

Q: 54 t-tests are used to test a sample mean when the population mean is unknown.

Q: 53 When the population standard deviation is known, z scores are appropriate to test a sample mean.

Q: 52 The standard deviation of a sampling distribution is known as the standard error.

Q: 51 A normal distribution is one in which all outcomes are equally likely.

Q: 50 A sampling distribution of the mean is typically the mean of one sample.

Q: 49 According to the Central Limit Theorem, as the number of samples increases, the distribution will approach the normal distribution.

Q: 48+ The term "effect size" refers to a) how large the resulting t statistic is. b) the size of the p value, or probability associated with that t. c) the actual magnitude of the mean or difference between means. d) the value of the null hypothesis.

Q: 47 The confidence intervals for two separate samples would be expected to differ because a) the sample means differ. b) the sample standard deviations differ. c) the sample sizes differ. d) all of the above

Q: 46 Thet distribution a) is smoother than the normal distribution. b) is quite different from the normal distribution. c) approaches the normal distribution as its degrees of freedom increase. d) is necessary when we know the population standard deviation.

Q: 45 If we compute a confidence interval as 12.65 25.65, then we can conclude thata) the probability is .95 that the true mean falls between 12.65 and 25.65.b) 95% of the intervals we calculate will bracket c) the population mean is greater than 12.65.d) the sample mean is a very precise estimate of the population mean.

Q: 44 A confidence interval computed for the mean of a single sample a) defines clearly where the population mean falls. b) is not as good as a test of some hypothesis. c) does not help us decide if there is a significant effect. d) is associated with a probability statement about the location of a population mean.

Q: 43 When would you NOT use a standardized measure of effect size? a) when the difference in means is itself meaningful b) when it is clearer to the reader to talk about a percentage c) when some other measure conveys more useful information d) all of the above

Q: 42 When you have a single sample and want to compute an effect size measure, the most appropriate denominator is a) the variance of the sample. b) the standard deviation of the sample. c) the sample size. d) none of the above