Deck 10: The Simple Experiment

Question 1

Internal validity refers to the soundness of the conclusion that

A) manipulating the independent variable caused a change in scores on the dependent variable
B) the results can be generalized safely to other species
C) the results can be generalized safely to other environments
D) the study accurately measured and manipulated the relevant variables

Answer 1

manipulating the independent variable caused a change in scores on the dependent variable

Question 2

The research method with the greatest internal validity is

A) field observation
B) the survey method
C) nonexperimental research methods
D) the simple experiment

Answer 2

the simple experiment

Question 3

The purpose of the simple experiment is to

A) describe behavior
B) predict behavior
C) explain behavior
D) eliminate variability

Answer 3

explain behavior

Question 4

Which of the following is NOT a prerequisite for establishing causality with the simple experiment?

A) statistical tests
B) eliminating random error
C) accounting for random error
D) independent random assignment

Question 5

The key to the simple experiment is

A) control over extraneous variables
B) eliminating extraneous variables
C) random assignment
D) keeping all nontreatment factors constant

Question 6

If the average score for the treatment group is higher than the average score for the no-treatment group, this difference could be due to

A) the treatment
B) random error
C) both a and b
D) neither a nor b

Question 7

If the difference between the treatment group and the no-treatment group is NOT larger than what would be expected by chance, then

A) the results are statistically significant
B) the treatment did not have an effect
C) the results are inconclusive
D) a Type I error has been made

Question 8

When doing an experiment and analyzing the results, a scientist hopes to

A) prove the null hypothesis
B) prove the experimental hypothesis
C) reject the null hypothesis
D) both b and c

Question 9

The relationship between the independent and dependent variables that you would like to demonstrate

A) is called the null hypothesis
B) is call the experimental hypothesis
C) requires descriptive statistics
D) requires correlational statistics

Question 10

The null hypothesis

A) supports the experimental hypothesis
B) contradicts the experimental hypothesis
C) reflects the true state of affairs
D) reflects the truth 95% of the time

Question 11

In an experiment to determine whether TV cartoons produce more aggressive behavior in grade school children, the null hypothesis states that TV cartoons

A) have an effect on aggressive behavior
B) have at least a slight effect on aggressive behavior
C) have no effect on aggressive behavior
D) have at most a weak effect on aggressive behavior

Question 12

In an experiment, the variable that is manipulated is the

A) dependent variable
B) participant
C) independent variable
D) extraneous variable

Question 13

The dependent variable in an experiment is

A) directly manipulated by the investigator
B) a variable whose effect is being studied
C) a variable that may influenced by the treatment
D) both a and b

Question 14

In an experiment on the effects of lighting on happiness, happiness would be the

A) control variable
B) independent variable
C) experimental condition
D) dependent variable

Question 15

Independent and dependent variables correspond to

A) internal and external validity, respectively
B) external and internal validity, respectively
C) causes and effects, respectively
D) effects and causes, respectively

Question 16

Independent is to dependent as ______ is to ______.

A) measured; manipulated
B) manipulated; held constant
C) held constant; measured
D) manipulated; measured

Question 17

Which of the following is NOT true of the control group?

A) it is a group of interacting individuals
B) participants are assigned to it via independent random assignment
C) it is compared with the experimental group
D) the more participants in the control group, the more closely it will reflect the experimental group - if the null hypothesis is correct

Question 18

The control group

A) has all its behavior under the experimenter's control
B) isn't measured on the dependent variable
C) doesn't get the treatment
D) is more standardized than the experimental group

Question 19

The participants who receive the treatment are called

A) independent variables
B) dependent variables
C) the control group
D) the experimental group

Question 20

Which of the following would be a drawback to running all members of the experimental group in a single morning session and all members of the control group in a single afternoon session?

A) one participant's responses may affect other participants' responses
B) differences between the control and experimental group might be due to time of day rather than to the treatment
C) chance differences between testing sessions would not have an opportunity to even out
D) all of the above

Question 21

Which of the following could be an experiment?

A) a study designed to determine whether men are more likely to smoke than women
B) a study that compares factory workers to managers
C) a study that randomly assigns some participants to judge a female job applicant's resume and others to judge a male job applicant's resume.
D) a study that compares a random sample of depressed people to a random sample of non-depressed individuals

Question 22

A statistically significant difference between two sample groups is NOT likely to be

A) a reflection of true differences between the populations they represent
B) due to chance
C) observed more than 5 percent of the time the groups are compared
D) observed when the two groups are very large

Question 23

What does it mean when a psychologist describes the results of an experiment by saying "The difference between groups is statistically significant"?

A) the difference between groups is very large
B) there is an important difference between groups
C) the observed difference is not likely to be due to chance
D) the observed difference is due to chance

Question 24

When using a statistical test to help us determine whether something occurred on the basis of luck alone, we would reject chance if the test suggested we had

A) a statistically significant finding
B) controlled for extraneous variables
C) used the independent variables effectively
D) a random event

Question 25

A statistical test can help us to determine whether or not a research finding was due to chance. When the test says that the finding was statistically significant, this means that

A) chance caused the finding
B) the factor under study had no effect
C) chance is an unlikely explanation for the results
D) unusual circumstances did not account for the finding

Question 26

In a research study, the null hypothesis is rejected when an analysis of the data suggests that the factor under study produced an effect that was

A) correlated with other factors under study
B) due to chance
C) statistically significant
D) consistent with the hypothesis under study

Question 27

Which of the following statements most accurately describes what assertion is being made when you declare the results of an experiment to be statistically significant?

A) you are certain that the independent variable had a large effect
B) your results have social or theoretical importance
C) you are confident that the null hypothesis is true
D) you expect that if the experiment was repeated, you would get a similar pattern of results

Question 28

If your results are not significant, but in the predicted direction, you can conclude

A) that the null hypothesis should be rejected
B) that the null hypothesis should be accepted
C) that the treatment had an effect, but this effect didn't reach statistical significance
D) none of the above

Question 29

If we reject the null hypothesis at the .05 probability level, then the differences we have found are such that they occur by chance

A) more than 95 out of 100 times if the null hypothesis were true
B) less than 5 out of 100 times if the null hypothesis were true
C) more than 95 out of 100 times if the null hypothesis were false
D) less than 5 out of 100 times if the null hypothesis were false

Question 30

The results of two experiments were analyzed. The results of Experiment A were significant at the .05 level and the results of Experiment B were significant at the .01 level of confidence. We can be more confident that the null hypothesis is false in

A) Experiment A
B) Experiment B
C) both experiments since the results were significant
D) neither experiment since the probability levels are so low

Question 31

If the null hypothesis is false and we reject the null hypothesis, we have made a

A) correct decision
B) mistake
C) type I error
D) type II error

Question 32

If the level of significance equals .05 and the null hypothesis is true, a correct decision will occur

A) about 95% of the time
B) about 70% of the time
C) about 5% of the time
D) all of the time

Question 33

The null hypothesis presumes that any observed difference between sample means

A) is a probable outcome
B) reflects an effect
C) is due to chance
D) occurs rarely

Question 34

Experimenter A and Experimenter B do essentially the same experiment, but when analyzing the results, Experimenter A uses a .05 level of significance, whereas Experimenter B uses a .01 level of significance. Who is most likely to make a Type 2 error?

A) Experimenter A
B) Experimenter B
C) both experimenters are equally likely
D) can't tell from information given

Question 35

Experimenter A and Experimenter B do essentially the same experiment, but when analyzing the results, Experimenter A uses a .05 level of significance, whereas Experimenter B uses a .01 level of significance. Who is most likely to make a Type 1 error?

A) Experimenter A
B) Experimenter B
C) both experimenters are equally likely
D) can't tell from information given

Question 36

An experiment may produce null results because

A) the levels of the independent variable are too similar to each other
B) not enough participants are used held
C) the study was not sufficiently standardized
D) all of the above

Question 37

As the sample size increases:

A) the power of a test increases
B) the probability of a Type II error decreases
C) both a and b
D) neither a nor b

Question 38

Which of the following will decrease the chances of making a Type 1 error?

A) increasing sample size
B) using a reliable measure of the dependent variable
C) using a homogenous group of participants
D) none of the above

Question 39

Which of the following will decrease the chances of making a Type 2 error?

A) increasing sample size
B) using a reliable measure of the dependent variable
C) using a homogenous group of participants
D) all of the above

Question 40

While observing an experiment in the laboratory, you note that neither the researcher interacting with the participant nor the participant knows which group the participant is in. You are observing a

A) single-blind experiment
B) double-blind experiment
C) poorly designed experiment
D) Frankenstein movie

Question 41

If neither the experimenter nor the participant knows if the drug the participant is taking is a placebo or "the real thing,"this method of eliminating experimenter bias is called:

A) pilot study
B) standardization
C) single blind
D) double blind

Question 42

Using an empty control group may

A) improve power
B) hurt construct validity
C) cause interpretation problems that could be solved by using a placebo treatment
D) all of the above

Question 43

________ statistics are used to draw conclusions from a set of scores.

A) Descriptive
B) Inferential
C) Correlational
D) Experimental

Question 44

We use inferential statistics to

A) help us decide whether sample observations hold true for populations
B) help us decide whether population observations hold true for samples
C) help us describe our data
D) organize and summarize a set of scores

Question 45

Inferential statistics can help us to correctly generalize from a sample to a population.

A) most of the time, if the sample is a random sample
B) always, if the sample is a random sample
C) always, if the sample is large enough
D) most of the time, if the sample is large enough

Question 46

If you found that the mean score of your experimental group differed from the mean score of your control group, then you could conclude that the difference

A) was due to the experimental treatment
B) was due to sampling or measurement error
C) was significant
D) none of the above

Question 47

In order to decide whether observed differences between samples reflect actual differences between populations, it is necessary to determine the _________ of the observed differences.

A) within groups variability
B) mean
C) absolute size
D) statistical significance

Question 48

An observed difference between two groups is MORE likely to be statistically significant if

A) the observed difference between the means is large
B) the observed differences between the scores within each group is small
C) the groups are large
D) all of the above are true

Question 49

Suppose that after statistical analysis, you decided that the scores of your experimental group were from a different population that the scores of your control participants. You would then have to conclude that

A) the experiment was useless, since there is no way to tell which two populations your samples represent
B) serious confounding has occurred, and the experiment must be repeated more carefully
C) there was insufficient evidence to reject the null hypothesis
D) your experimental treatment probably had an effect

Question 50

You conduct a simple experiment. Your experimental group subjects are best viewed as

A) a random sample from a very large population
B) a unique population
C) a random sample of all the participants at your school
D) a random sample of all the participants who participated in your experiment

Question 51

According to the central limit theorem,

A) the standard deviation will always be smaller than the standard error
B) the standard deviation will always be larger than the standard error
C) the distribution of sample means will be normally distributed if your sample sizes are large enough
D) the distribution of sample means will be normally distributed if the underlying population is normally distributed

Question 52

A statistical test is most likely to indicate significant results when the difference between the means is _______ and the within-groups variability for each both the control and experimental groups is ______.

A) large; large
B) large; small
C) small; large
D) small; small

Question 53

In inferential statistics the chance of erroneous generalizations can be

A) eliminated
B) estimated
C) ignored
D) bypassed

Question 54

The t test for the simple experiment has degrees of freedom equal to the

A) the total number of participants
B) the total number of participants minus one
C) the total number of participants minus two
D) the total number of participants minus three

Question 55

The only way to establish whether one variable causes another is to conduct an experiment.

Question 56

Participant age is a an appropriate independent variable for an experiment.

Question 57

The prediction that the treatment will cause an effect is called the null hypothesis.

Question 58

If your treatment does not seem to have an effect then you accept the null hypothesis.

Question 59

The independent variable in an experiment is the suspected cause of a predicted outcome.

Question 60

In order to establish causality in a simple experiment, participants' scores must be independent.

Question 61

Independent random assignment is the cornerstone of a simple experiment because it removes the effects of chance differences between groups.

Question 62

If your study does not have independent random assignment, then your study cannot be an experiment.

Question 63

A statistically significant t test tells you that the findings are very important.

Question 64

Type 1 errors occur when you mistake a chance difference for a treatment effect.

Question 65

With a little forethought, you can easily design an experiment that reduces the chances of making a Type 1 error.

Question 66

The best way to improve the chances of making a Type 2 error is to increase your experiment's power.

Question 67

The t test is always the best way to analyze the results from a simple experiment.

Question 68

What are the assumptions of the t test? Why are these assumptions necessary?

Question 69

What questions are raised by a simple experiment that obtains null results?

Question 70

What questions are raised by a simple experiment that obtains significant results?

Question 71

Discuss the tradeoffs that a researcher may have to make when trying to increase power.