Deck 9: Experimental Design

Question 1

A researcher is interested in environmental effects on workers' productivity. Each of 20 call-centre workers is tested twice, one day in a noisy open-plan office and the next day in a quiet individual office. A record is made of how many calls they manage to process in each day. What is the major problem with this study?

A) The lack of counterbalancing means that order effects may affect the data.
B) Number of calls processed is a poor measure of performance.
C) Twenty call-centre workers is too few for a study like this.
D) The study hasn't identified exactly what sorts of environmental factors might have produced the results.

Answer 1

A

Question 2

A researcher looked at how aesthetic ratings were affected by beliefs about a painting's attributes. One group of art historians were shown three paintings, and told that the first was by Picasso, the second by Rembrandt and the third by Monet. Another group were shown the same paintings, but told they were by amateur artists. Each art historian was then asked to independently rate each painting for its aesthetic impact. This is an example of a:

A) matched pairs design, with two independent variables: the fame of the painter and the aesthetic impact of the painting.
B) mixed design, with two independent variables: the apparent fame of the painter, and the type of painting.
C) mixed design, with two independent variables: the type of painting, and its rated aesthetic impact.
D) an independent-measures design, with one independent variable: the type of painting.

Answer 2

B

Question 3

In the 18th and 19th centuries, phrenologists claimed that cognitive and emotional attributes were reflected in the shape of the skull (which is nonsense, as brain shape is unrelated to skull shape, and in any case brain size bears no relationship to ability). Measurements of skull shape within an individual are very consistent over time. This is an example of data that show:

A) reliability without validity.
B) validity without reliability.
C) poor reliability and validity.
D) low validity and poor generalisability.

Answer 3

A

Question 4

A sadistic pastry chef wants to test the effect of gingerbread house foreclosure on the stress level of his gingerbread people. He separates the gingerbread people randomly into two groups. The chef eats the gingerbread houses of one group AND for good measure eats the gumdrop buttons of each the gingerbread people while the other group is left entirely alone. What is a confounding variable in the pastry chef's experiment?

A) Eating the gingerbread house of only one group and not the other
B) The measured stress level of the gingerbread people
C) Eating the gumdrop buttons of only one group and not the other
D) There is no confounding variable in this experiment

Question 5

Reducing noise in data involves:

A) minimising random variation wherever possible.
B) adding error to a true score for each participant.
C) ensuring that all participants are quiet during experimental procedures.
D) comparing the true scores of one group with the true scores of another.

Question 6

A reflexochirosteopathic clinic performs a study to demonstrate the effectiveness of their treatment for back pain. They recruit 50 patients. After a month, ten remain. All ten report some improvement in their condition, which is hailed by the clinic as validating the efficacy of their treatment. Which of the following factors might be influencing these results?

A) Placebo effect (remaining participants' belief that the treatment has worked has affected their ratings of its effectiveness)
B) Differential mortality (participants for whom the treatment didn't work dropped out of the study)
C) Regression to the mean (participants joined the study when their pain was at its worst, and so it is likely to have decreased anyway)
D) All of these

Question 7

An experimenter is interested in the effects of watching TV on children's cognitive development. She recruits two groups of 20 boys, one containing children who regularly watch TV for two hours a day, and the other comprised of children who never watch TV. When tested, the groups are significantly different in problem-solving ability. Which of the following is a threat to the validity of these results?

A) The experimenter hasn't taken account of which channel the children watch.
B) The groups are initially non-equivalent.
C) Two hours isn't long enough to show any effects of TV watching.
D) The study should have included both boys and girls.

Question 8

Which of the following is NOT a way to reduce noise in data?

A) Using a within subjects design
B) Clearly operationalising independent and dependent variables
C) Maintaining consistent procedures throughout all conditions of an experiment
D) Increasing the number of experimental conditions

Question 9

A researcher is interested in whether eyewitnesses' performance will be affected by the culprit wearing a hat. Half of the participants will be shown a video of a crime where the suspect is wearing a hat, and will then see a lineup containing the suspect wearing a hat. The other half will see the same video, but the suspect will not be wearing a hat when they appear in the lineup. The researcher is worried that differences in visual acuity between the two groups of participants might be a confounding variable in this study. She can avoid this by:

A) dividing up participants into conditions based on visual acuity.
B) randomising the assignment of participants into her two conditions.
C) removing all conditions and testing memory only on the basis of visual acuity.
D) correlating visual acuity with memory ability.

Question 10

Identical (monozygotic) twins are recruited for a study of a new learning procedure to improve memory and factor out individual variation due to genetic differences. One from each pair of twins is randomly assigned to a learning procedure and the other is assigned to a control condition. This is a ______ design.

A) genetic
B) within-subjects
C) naturalistic
D) matched pairs

Question 11

Two friends, Tom and Harry, are bowling. Tom never manages to get a strike, hitting the same 3 pins every frame. Harry on the other hand sometimes hits a strike, sometimes a few pins, and sometimes none at all. It could be said that Tom, while lacking in accuracy, has ______ while Harry does not.

A) skill
B) validity
C) maturation
D) reliability

Question 12

A health psychologist wants to test a behavioural intervention for helping adult smokers quit smoking. He decides to test his intervention on 9-year-olds who have never smoked real cigarettes and are instead addicted to sweet cigarettes. Even if the intervention is shown to be effective in helping 9-year-olds kick their sweet cigarette habit, this study might still be criticised for lacking ______.

A) scrutiny
B) internal validity
C) external/ecological validity
D) demand characteristics

Question 13

A researcher wanted to test whether people tend to use less smartphone technology. as they get older. The researcher surveyed 10-year-olds, 30-year-olds, 50-year-olds, and 70-year-olds, and found that the 10-year-olds and 30-year-olds used smartphones more than the older respondents. The researcher concluded that cognitive ageing causes people to use their phone less. What is a concern for this conclusion?

A) The sample is biased because there are no people between 21‒29 years old.
B) There may be cohort effects.
C) The Hawthorne effect.
D) The age range is too large to make comparisons.

Question 14

A teacher gave his class a brief quiz to assess their knowledge. On the basis of this test, he gave extra tuition to the pupils who finished in the lower 50% of the class. After one month, he gave them another quiz. He found that the pupils who received the extra tuition improved more than the other pupils. Why might the teacher have observed this apparent difference in improvement?

A) Testing effect
B) Crossover effects
C) Differential mortality
D) Regression towards to the mean

Question 15

A researcher is interested in how well a new anti-depressant drug works for people with mild depression. She recruits a group of 120 people with mild depression, measures their depression, and randomly allocates them into three conditions. The first condition receives a treatment of the drug for two weeks. The second condition believes that they are taking the drug for two weeks, but in fact they are given sugar pills. The third condition is a control group who are given nothing to take. After two weeks, the first two groups each show large improvements of a similar magnitude, while the third group show no improvement. The results of this study suggest a strong:

A) effect of the anti-depressant drug.
B) placebo effect.
C) effect of evaluation apprehension.
D) carryover effect.

Question 16

The Pygmalion Effect refers to:

A) when an experimenter's beliefs about participants affect their performance.
B) when a confound in a Latin Square design creates an effect.
C) when a confound in a Solomon four-group design creates an effect.
D) when participants are not randomly allocated to the control group.

Question 17

Which of the following is/are an advantage of longitudinal designs over cross-sectional designs?

A) The power to detect an effect is usually higher for the same number of participants.
B) They prevent cohort effects.
C) They allow for estimates of how individuals change.
D) All of these.

Question 18

Why might a researcher conduct a multifactorial design rather than one with a single independent variable?

A) It allows the researcher to use a smaller number of participants.
B) It creates simpler statistics for complex dependent variables.
C) It allows the researcher to see if the effect of one independent variable depends on the level of another independent variable.
D) All of these.

Question 19

Suppose a researcher was interested in effect of milk on children's height. The researcher recruited 20,000 schoolchildren and measured their height. They randomly allocated pupils to either receive milk or no milk, but they told teachers that if it appeared that if either the milk or the no milk groups in their classroom appeared to have more well-nourished children, they could even the groups out. After the experiment it was found that the children in the milk condition had grown more (but their average height was still less than the no milk group). Which of the following is the biggest problem with this study?

A) The sample size is too large.
B) There was a bias in the allocation of participants to conditions.
C) The group that gains the most height must be taller, there were problems in the calculations.
D) The placebo effect.

Question 20

A study like that described in the previous question was actually conducted. The statistician William Gossett (who used the name Student when he devised the t-test, discussed in later chapters), described how instead of randomly allocating 20,000 pupils into conditions, the researchers could have just used a couple of dozen pairs of monozygotic (identical) twins, randomly allocating one in each pair to each condition. This is called a:

A) matched-pairs design.
B) Latin squares design.
C) between-subjects design.
D) cross-sectional design.