Deck 11: Expanding the Simple Experiment: The Multiple Group Experiment

Question 1

The random assignment of subjects to two or more experimental conditions produces groups that are

A) equal at the outset of the experiment
B) similar at the outset of the experiment
C) noticeably different at the outset of the experiment
D) noticeably different at the end of the experiment

Answer 1

similar at the outset of the experiment

Question 2

Which of the following is an advantage of using a multiple group experiment over a simple experiment?

A) You can compare more than two kinds of treatments.
B) You can use more than one control group.
C) You can determine the shape of the functional relationship among variables.
D) all of the above

Answer 2

all of the above

Question 3

The advantage of comparing two kinds of treatment with a no-treatment group is

A) we can do ANOVA
B) we can do an F test
C) we can determine whether the treatments have a positive or negative effect
D) all of the above

Answer 3

we can determine whether the treatments have a positive or negative effect

Question 4

Using multiple levels of a quantitative independent variable can

A) increase your ability to accurately estimate the effects of unexplored levels of the independent variable
B) help alert you to problems caused by an empty control group
C) decrease confounding variables
D) all of the above

Question 5

Using more than two levels of an independent variable can

A) increase your study's external validity
B) increase the chances that your study will find a significant effect
C) both a and b
D) neither a nor b

Question 6

If people perform poorly when not motivated, well when moderately motivated, and poorly when highly motivated, the relationship between motivation and performance is

A) linear
B) "u"-shaped
C) double "u"-shaped
D) functionally dubious

Question 7

To test to see if the functional relationship between two variables is "u"-shaped, you need at least _____ levels of the independent variable.

A) one
B) two
C) three
D) four

Question 8

If we do not know the true functional relationship between the independent and dependent variable, then

A) our results are not statistically significant
B) we cannot accurately generalize our results to unexplored levels of the independent variable
C) our study will lack construct validity
D) our study will lack internal validity

Question 9

According to the text, to pin down the functional relationship between variables you will usually need _____ levels of the independent variable.

A) two
B) three
C) four
D) five

Question 10

Random assignment

A) allows us to establish internal validity
B) allows us to rule out the effects of every variable except for the independent variable
C) both a and b
D) neither a nor b

Question 11

Which would be the least useful advice for ruling out the effects of confounding variables

A) pay close attention to how you treat your experimental group
B) avoid empty control groups
C) use placebos
D) use several control groups

Question 12

The construct validity of a simple experiment should be questioned, especially if it used

A) random assignment
B) matching
C) an empty control group
D) fewer than 30 participants

Question 13

Within group variability is caused by which of the following?

A) individual differences
B) chance factors
C) the independent variable
D) a and b

Question 14

The reason why a particular participant's score was not the same as the other participants' average score (all of whom received the same treatment) is due to

A) variation caused by the IV
B) the mood of the subject
C) unidentified or unmeasured chance factors
D) bad research technique

Question 15

If the null hypothesis is false, variability between groups reflects

A) the treatment effect
B) random error
C) random error and possibly the treatment effect
D) random error and the treatment effect

Question 16

If the null hypothesis is true, variability between groups reflects

A) the treatment effect
B) random error
C) random error and possibly the treatment effect
D) random error and the treatment effect

Question 17

Suppose you are analyzing the results from a study. When you first see the variability between groups, you should probably view that variability as representing

A) the treatment effect
B) random error
C) random error and possibly the treatment effect
D) random error and the treatment effect

Question 18

Regardless of whether the null hypothesis is true or false, variability within groups reflects

A) the treatment effect
B) random error
C) random error and possibly the treatment effect
D) random error and the treatment effect

Question 19

In analysis of variance, differences between group means are represented by ____ variability.

A) between groups
B) between subjects
C) within groups
D) within subjects

Question 20

An F test is based on the idea that, if the null hypothesis is true, the top part of the F ratio tends to be

A) substantially larger than the bottom part
B) larger than the bottom part
C) smaller than the bottom part
D) about the same as the bottom

Question 21

Use analysis of variance rather than a t test whenever you are dealing with

A) more than one group mean
B) more than two group means
C) an empty control group
D) normal distributions

Question 22

Which of the following would NOT be a source of within group variability?

A) treatment effects
B) individual differences
C) unreliability of the dependent measure
D) lack of standardization

Question 23

Which of the following would be a source of between group variability?

A) treatment effects
B) individual differences
C) both a and b
D) neither a nor b

Question 24

______ is the same as ________ .

A) MS Between, MS Error
B) MS Error, MS Within
C) MS Treatment, MS Within
D) MS Treatment, MS Error

Question 25

The F test would be preferable to the t test for which of the following designs?

A) two-group design
B) three-group design
C) both a and b
D) neither a nor b

Question 26

Using a t test rather than an F test with a five group experiment would

A) increase your chances of making a Type 1 error
B) decrease your chances of making a Type 1 error
C) decrease your chances of making a Type 2 error
D) decrease your power significantly

Question 27

If an investigator explained the statistical analysis she performed on a design by saying she was dividing MS Between by MS Within, what, in fact, was she saying?

A) that she was calculating the basic F test
B) that she was performing an analysis of covariance statistical analysis
C) that she was calculating a post hoc test
D) that she was didn't know what she was doing

Question 28

If an investigator explained the statistical analysis she performed on a design by saying she was dividing SS Between by SS Within, what, in fact, was she saying?

A) that she was calculating the basic F test
B) that she was performing an analysis of covariance statistical analysis
C) that she was calculating a post hoc test
D) that she was didn't know what she was doing

Question 29

If only chance is operating to influence your data, then the theoretically expected F ratio would be

A) around 0.00
B) around 1.00
C) negative
D) around 20.00

Question 30

If a treatment effect exists

A) variability between groups will tend to exceed variability within groups
B) variability within groups will tend to exceed variability between groups
C) variability between groups and within groups will tend to be about the same
D) total variability will tend to exceed variability between groups and within groups

Question 31

Essentially, the F ratio reflects the ratio of the

A) true null hypothesis to the false null hypothesis
B) treatment effect to the null hypothesis
C) observed sample mean differences to the null hypothesis
D) observed sample mean differences to the estimated error term

Question 32

In analysis of variance, indexes of variability within and between groups consist of the

A) differences between sample means divided by the degrees of freedom
B) differences between sample means divided by the sample size
C) sum of squares divided by the degrees of freedom
D) sum of squares divided by the sample size

Question 33

In analysis of variance, the index of variability within groups and the index of the variability between groups are often referred to as

A) variance means
B) mean squares
C) F tests
D) F ratios

Question 34

In a table that summarizes the results of an analysis of variance, the word "Between"might be replaced by

A) "Within"
B) "Error"
C) "Treatment"
D) "Total"

Question 35

In a table that summarizes the results of an analysis of variance, the word "Within"might be replaced by

A) "Between"
B) "Error"
C) "Treatment"
D) "Total"

Question 36

What does the "F score"for the treatment main effect tell the researcher?

A) If F is larger than 1 then the IV had an effect.
B) The F ratio indicates which group significantly differed from the other.
C) It tells her nothing until she can see if the F she calculated would occur by chance in like conditions less than 5% of the time by looking it up on an F table.
D) a and b

Question 37

The "critical F"found in a table labeled "Critical Values of F for the .05 Level of Significance"is

A) the true answer to the problem
B) the F that would be obtained or exceeded 5% of the time by chance alone
C) the F score which is due to 95% chance
D) the point at which significance ceases

Question 38

If you are conducting an ANOVA and you have 7 degrees of freedom between groups, the number of groups in the study equals

A) 6
B) 7
C) 8
D) cannot be answered on the basis of information given

Question 39

In an experiment involving four different groups, each consisting of 5 participants, the degrees of freedom for between groups equals

A) 3
B) 4
C) 5
D) 20

Question 40

If the calculated F score obtained in an analysis of a three-group design is larger than the "critical F,"the investigator may conclude which two of the following?

A) The F value she calculated would occur by chance more than 5% of the time.
B) The group means are all significantly different.
C) The F value she calculated would occur by chance less than 5% of the time.
D) The group means are not significantly different.

Question 41

In an ANOVA, rejection of the overall null hypothesis indicates that

A) at least two population means differ
B) at least several population means differ from each other
C) at least some population means are equal
D) all population means differ from each other

Question 42

To pinpoint specific differences between pairs of population means,

A) use a t test
B) use a post hoc test
C) use a F test
D) inspect the observed differences between sample means

Question 43

Which of the following conditions must you meet to do a post hoc trend analysis?

A) obtain a statistically significant overall effect
B) have a quantitative independent variable
C) have a dependent measure that yields interval or ratio scale data
D) all of the above

Question 44

If you have 3 levels of the independent variable, your trend analysis can reveal

A) one trend
B) two trends
C) three trends
D) four trends

Question 45

In an experiment involving four different groups, each consisting of 5 participants, the total degrees of freedom equals

A) 16
B) 18
C) 19
D) 20

Question 46

In an experiment involving four different groups, each consisting of 5 subjects, the degrees of freedom for your error term equals

A) 16
B) 18
C) 19
D) 20

Question 47

In an experiment involving four different groups, each consisting of 5 subjects, the degrees of freedom for your treatment term equals

A) 3
B) 4
C) 5
D) 20

Question 48

The multiple-group experiment is more sensitive to nonlinear relationships than the simple experiment.

Question 49

In order to use the multiple-group experiment to discover the functional relationship, your dependent measure must provide at least ratio scale data.

Question 50

Simple experiments usually have more construct validity than multiple-group experiments.

Question 51

The most common way to first analyze data from a multiple-group experiment is to compute an analysis of variance (ANOVA).

Question 52

The only reason treatment groups in a multiple-group experiment will differ from one another is due to differences in their treatments.

Question 53

Scores within a treatment group differ from each other due to random error.

Question 54

The F test is designed to see whether the differences between the group means are greater than would be expected by chance.

Question 55

If you conducted an experiment with 6 groups and 50 participants, you would look in a F table under 6 and 44 degrees of freedom.

Question 56

You are most likely to get a significant F if the between-groups variability is small.

Question 57

If you get a significant F in a five-group experiment, you should conduct post hoc tests to discover which groups are reliably different.