PSYC 60 Final Review

If α increases, which of the following will NOT happen?

• The probability of incorrectly rejecting H0 will increase

• The probability of incorrectly retaining H0 will increase

• The probability of incorrectly retaining H0 will decrease

• The probability of correctly rejecting H0 will increase

If variability is ____, scores are more spread out and less clustered together.

• No,  χ2 = 0.332

• Yes,  χ2 = 0.505

• No,  χ2 = 0.566

• No,  χ2 = 0.668

Answer explanation

If a researcher is interested in which UCSD lecture halls students have the most lectures in: what type of variable are they measuring, what type of frequency distribution should be used, and what measure of central tendency is most appropriate?

• nominal, histogram, median

• nominal, bar graph, mode

• ordinal, bar graph, median

• ordinal, polygon, mode

Answer explanation

Lecture halls are categories/names, and they don't have a particular order (Peterson doesn't necessarily come before Catalyst) --> this is nominal data. Recall that only bar graphs can be used for qualitative/categorical data (Chapter 2). The mode is the most appropriate measure of central tendency for nominal data (Chapter 3) :).

A normal distribution of scores has a mean of μ = 20 and a standard deviation of σ = 4. What is the probability of randomly selecting a score greater than 26?

• 0.3215

• 0.9332

• 0.6785

• .0668

Answer explanation

Chapter 6: The question does not include information about the sample size (n) its wording points to scores, rather than sample means. So, we're working with a distribution of single scores, rather than sample means (Chapter 7).

We're given a raw score (X) and are expected to give a proportion back (a "forwards" probability question)

Which of the following will cause a researcher the most problems when trying to demonstrate statistical significance using a two-tailed independent-measures t-test?

• A. High variance

• B. Low variance

• C. Large/high sample means

• D. Small/low sample means

Answer explanation

(Chapter 10)

A. High variance will increase our standard error (the denominator in the test statistic) and decrease the test statistic overall, making it harder to reject the null hypothesis

B. Low variance will decrease standard error and increase our test statistic, making it easier to reject the null hypothesis)

C & D. The sample means being large or small values won't affect your test statistic for independent measures because the numerator is measuring the difference between them - ex. a test statistic calculated from an M1 = 40 and M2 = 35 will have the same numerator as one calculated from a M1= 10 and M2 = 5.

A researcher is interested in the effects of a new drug on reaction time. The average reaction time is μ = 200 ms with σ = 25 ms in the general (normally distributed) population. The researcher plans to take a sample of n = 25 and predicts a 10 ms decrease due to the drug. If they use a one-tailed test with 𝛼 = 0.05, what would be the power of the test?

• power = 0.95

• power = 0.05

• power = 0.3632

• power = 0.6368

Answer explanation

• Ackerman and Goldsmith (2011) report that students who study from a screen (phone, tablet, or computer) tended to have lower quiz scores than students who studied the same material from printed pages. To test this finding, a professor identifies a sample of n = 16 students who used the electronic version of the course textbook and determines that this sample had an average score of M = 72.5 on the final exam. During the previous three years, the final exam scores for the general population of students taking the course averaged μ = 77 with a standard deviation of σ = 8 and formed a roughly normal distribution. The professor would like to use the sample to determine whether students studying from an electronic screen had exam scores that are significantly different from those for the general population. Use a two-tailed test with  α = .05.

Select the value of the appropriate test statistic and decision.

• -2.25, Fail to reject H0

• -2.25, Reject H0

• 2.25, Fail to Reject H0

• 2.25, Reject H0

Answer explanation

If given df = 19, which of the following must be true?

• For an independent measures t-test, there are 21 participants

• For a repeated measures t-test, there are 20 participants

• For correlation, there are 20 participants

• For a matched-subjects t-test, there are 20 participants

• For chi-squared goodness of fit, we don't know how many participants there are with just the df.

Answer explanation

Independent measures t-test: df = (n1-1) + (n2-1)

19 = (n1-1) + (n2-1)

19 + 2 = n1 + n2 --> 21

Repeated measures: df = n-1 (where n is the number of participants)

19 = n-1 --> 20

Correlation: df = n-2

19 = n-2 --> 21

Matched subjects: df = n-1 (where n is the number of difference scores - i.e. pairs of participants)

19 = n-1 --> 20

20 x 2 (to unpair) = 40

Chi-Squared GOF: df = C-1 (where C is the number of categories)

We don't know n with just the df!

If a sample has a mean of M = 10 and a standard deviation of s = 5, what will happen to M and s if 5 is added to every single score in the sample?

• M and s will not change

• M = 10, s = 10

• M = 15, s = 10

• M = 15, s = 5

Answer explanation

(Chapter 3) Adding/subtracting a constant to each score will +/- the same value from the mean

M = 10 + 5 = 15

(Chapter 4) Adding/subtracting a constant to each score will not change the standard deviation

s = 5