Deck 22: Inferential Data Analysis

Question 1

A method for simplifying the process of drawing conclusions from different studies on the same topic is called super-analysis.

Answer 1

False

Question 2

Inferential statistics are used primarily for describing samples.

Answer 2

False

Question 3

When two groups are significantly different on a dependent variable at the .05 level of significance, it means that there is afive-percent difference between them.

Answer 3

False

Question 4

The selection of the significance level​

A) is fixed at .05 for social science research.​
B) ​depends on the size of the population.
C) ​is arbitrary, although .05 is frequently used.
D) ​must be decided after finishing the analysis.

Question 5

The ultimate aim of every significance test is to discern the probability that an observed relationship between variables can be attributed to chance.

Question 6

Which of the following hypotheses assumes that there is no relationship between two variables?​

A) ​Research
B) ​Null
C) ​Alternate
D) ​Secondary
E) ​Substantive

Question 7

The smaller the sample, the greater the probability of obtainingstatistically significant findings.

Question 8

A correlation of .60 indicates that 36 percent of the variance in the dependent variable has been explained.

Question 9

We commit a Type II error when we​

A) ​accept a false null hypothesis.
B) ​accept a true null hypothesis.
C) ​reject a false null hypothesis.
D) ​reject a true null hypothesis

Question 10

If a result is statistically significant, it is also substantivelysignificant.

Question 11

A .05 level of significance means that​

A) ​there is only a 5% chance that the statistic's value could be obtained as a result of sampling errors only.
B) ​one is 50% certain that the sample value is representative of the population.
C) ​there is only a 5% chance that the variables tested are not independent.
D) ​the results can be accepted because the sampling error is only 5%.
E) the level of confidence is only 5%.​

Question 12

The more statistically significant a finding, the less likely it isto be explained away by extraneous variables or design flaws.

Question 13

If a Cohen's d effect size exceeds .6, the finding is statistically significant.

Question 14

The term "clinical significance" means the same thing as "substantive significance.

Question 15

A relationship with a correlation of .50​

A) ​has more substantive significance than any other relationship with a lower correlation.
B) ​shows that 50% of the variation in the dependent variable is explained.
C) ​is stronger than any relationship with a correlation of -.50.
D) ​is more likely to be statistically significant than any relationship with a lower correlation.
E) might not be substantively significant​

Question 16

When using inferential statistics, we must risk either a Type I error or a Type II error, we cannot avoid risking both.

Question 17

Statistical power analysis deals with the probability of avoiding Type I errors.

Question 18

Two-tailed tests of significance should be used whenever there isa non-directional hypothesis.

Question 19

A research article reports that 4 of 10 students who received an intervention from a school social worker dropped out of school, while 6 of 10 students in the control group dropped out. The difference was not statistically significant. What can be inferred from these statistics?​

A) ​The social worker's intervention was not effective.
B) ​The social worker's intervention was effective, but just not at a statistically significant level.
C) ​The same difference in drop out rates with a much larger sample might be statistically significant.
D) ​The same difference in drop out rates with a smaller sample might be statistically significant.

Question 20

In reference to sampling error, which of the following statements is valid?​

A) ​The smaller the sample, the less sampling error.
B) ​Sampling error is an inverse function of sample size.
C) ​Inferential statistics are addressed to sampling and non-sampling errors.
D) ​Random assignment ensures no sampling error.

Question 21

When a finding is statistically significant:​

A) ​alternative explanations other than the independent variable are not plausible.
B) ​a strong relationship has been observed.
C) ​it is an important finding.
D) ​its meaning depends on the rigor of the research methods.

Question 22

Explain what is meant by "chance as a rival hypothesis.

Question 23

In an evaluation of an intervention that seeks to reduce depression scores among the frail elderly, the mean of the experimental group is 30, the mean of the control group is 60, and the standard deviation is 10. Cohen's d effect size is​

A) ​3
B) ​minus 3
C) ​30
D) ​10

Question 24

A research article reports that clients who received Intervention A had a 20% recidivism rate, while clients who received Intervention B had a 30% recidivism rate. What can be inferred from these statistics?​

A) ​Intervention A is more effective than Intervention B.
B) ​The difference in recidivism rates may have been due to chance.
C) ​We risk a Type II error if we conclude that Intervention A is more effective than Intervention B.
D) ​We risk a Type I error if we conclude that the difference in recidivism rates may have been due to chance.

Question 25

Explain Type I and Type 11 errors, when they are risked, and what can be done to reduce their risks.

Question 26

An experiment is conducted to test the effectiveness of a social work intervention. The experimental group improves more than does the control group, but the results are not statistically significant. This means that​

A) ​the intervention was not effective.
B) ​the intervention was effective, but not at a significant level.
C) ​we cannot rule out sampling error in inferring about the effectiveness of the intervention.
D) ​the results were clinically significant despite not being statistically significant.

Question 27

A researcher conducts a test of statistical significance and finds a probability of less than .05. The researcher should NOT conclude that​

A) ​it is improbable that the observed relationship could have resulted from sampling error alone.
B) ​the null hypothesis was rejected.
C) ​there is a statistically significant relationship at the .05 level.
D) ​the results have important implications for social work.

Question 28

A Cohen's d effect size of .60 is

A) ​stronger than the average Cohen's d effect size reported in meta-analyses in social work and allied fields.
B) weaker than the average Cohen's d effect size reported in meta-analyses in social work and allied fields.​
C)about the same as the average Cohen's d effect size reported in meta-analyses in social work and allied fields.
D) ​substantively significant.
E) ​not substantively significant.

Question 29

Explain the difference between statistical significance, relationship strength, and substantive significance.

Question 30

How would you interpret a finding from an experiment in which a costly and time-consuming intervention to reduce school dropout had a statistically significant outcome with a Cohen's d effect size of .20? What if the effect size were .80, but the outcome were not statistically significant due to a low sample size?