Deck 10: Nonexperimental Research Descriptive and Causal-Comparative Designs

Question 1

In educational research, description

A) is a major goal of both quantitative and qualitative research.
B) rarely is the focus of data-collection efforts.
C) is associated with nonparametric tests of statistical significance.
D) focuses on demonstrating cause-effect connections.

Answer 1

is a major goal of both quantitative and qualitative research.

Question 2

Causal-comparative research differs most distinctly from descriptive research in its

A) use of measures of central tendency and measures of variability.
B) appropriateness for exploring how variables are related to each other.
C) potential for impact on educational practice.
D) study of the same population at different times.

Answer 2

appropriateness for exploring how variables are related to each other.

Question 3

Descriptive research can focus on depicting

A) educators' attitudes.
B) educators' professional behavior.
C) educators' beliefs.
D) all of the above.

Answer 3

all of the above.

Question 4

A descriptive research study would be likely to report all but which of the following types of scores or statistics?

A) Standard deviation
B) Mean
C) Percentiles
D) Analysis of variance

Question 5

Derived scores are useful in descriptive research because they

A) identify the amount of discrepancy between actual and ideal behavior or values.
B) show how an individual's performance compares with that of a comparison group.
C) explain how test scores are generated by individuals' performance on the test.
D) do not require an understanding of the normal probability curve.

Question 6

An opinion poll is an example of which type of descriptive research?

A) Description of a sample at one point in time
B) Cross-sectional research
C) Longitudinal research
D) Case study research

Question 7

Longitudinal research designs are particularly useful for studying

A) processes involved in human development.
B) cross-cultural differences in beliefs and values.
C) possible causal relationships between different variables.
D) differences in the views of various members of a defined population at a given time.

Question 8

If a researcher studies the same population over time by repeatedly contacting the same sample of individuals, the researcher is conducting astudy.

A) trend
B) cohort
C) panel
D) cross-sectional

Question 9

Match each of the following types of longitudinal research study with the appropriate description of what audience is studied at each data-collection point.

-Trend study

A) A different sample from a population that changes over time
B) A different sample from the same population
C) The entire population
D) The same sample from the same population

Question 10

Match each of the following types of longitudinal research study with the appropriate description of what audience is studied at each data-collection point.

-Cohort study

A) A different sample from a population that changes over time
B) A different sample from the same population
C) The entire population
D) The same sample from the same population

Question 11

Match each of the following types of longitudinal research study with the appropriate description of what audience is studied at each data-collection point.

-Panel study

A) A different sample from a population that changes over time
B) A different sample from the same population
C) The entire population
D) The same sample from the same population

Question 12

Compared with a longitudinal study, an important advantage of a cross-sectional study is that

A) information is collected at several points in time.
B) the data can be obtained more rapidly.
C) developmental changes can be detected more accurately.
D) the data are not distorted by faulty recall.

Question 13

Suppose you wish to examine the changes in attitudes toward school that occur as students mature. You select a sample of elementary, middle, and high school students and measure their attitudes toward school. A weakness of your cross-sectional design for studying this research problem is that

A) this method cannot be used to explore relationships between variables.
B) this method cannot be used to explore developmental changes.
C) the sample of elementary, middle, and high school students may differ on other characteristics besides grade level.
D) data must be collected over an extended time period.

Question 14

Match each of the following research results with the type of interpretation of it that a researcher is likely to make about the relationship among females' athletic participation (X), engaging in irresponsible sexual activity (Y), and receiving supportive parenting (Z).

-Female students who participate in school athletic events are less likely to engage in irresponsible sexual activity.

A) Z causes both X and Y.
B) Y causes X.
C) X causes Y.

Question 15

Match each of the following research results with the type of interpretation of it that a researcher is likely to make about the relationship among females' athletic participation (X), engaging in irresponsible sexual activity (Y), and receiving supportive parenting (Z).

-Females with supportive parents are more likely to participate in athletic events while in school and to make responsible decisions about sexual activity.

A) Z causes both X and Y.
B) Y causes X.
C) X causes Y.

Question 16

Match each of the following research results with the type of interpretation of it that a researcher is likely to make about the relationship among females' athletic participation (X), engaging in irresponsible sexual activity (Y), and receiving supportive parenting (Z).

-Females who engage in irresponsible sexual activity have less time and interest for athletic participation while in school.

A) Z causes both X and Y.
B) Y causes X.
C) X causes Y.

Question 17

In causal-comparative research, matching is used to

A) equate two groups on one or more extraneous variables.
B) identify groups that represent the extremes of a phenomenon.
C) predict the direction of a presumed causal relationship.
D) ensure that groups with different characteristics are selected randomly.

Question 18

The best way to select two groups for comparison in a causal-comparative research study so that they are highly different on the presumed causal variable is

A) through the use of matching.
B) by random selection of the groups from the defined population.
C) by determining the range of scores of the defined population on a measure of the variable and then selecting individuals whose scores are at the extreme of the range.
D) by using theory to determine the hypothetically extreme ranges of the presumed causal variable and then recruit volunteers who satisfy the requirements of the these ranges.

Question 19

After the researchers have selected a group having the characteristic they wish to study in a causal comparative design, the next logical step is to

A) randomly form subgroups from the selected group.
B) identify a comparable group without the selected characteristic.
C) collect biographical data on this group.
D) choose the statistical analyses to be performed.

Question 20

The t test should be used instead of the z distribution to test the difference between two means when

A) the size of the population is difficult to estimate.
B) the scores of the two samples are correlated.
C) N is less than 30.
D) very large samples are studied.

Question 21

For which of the following hypotheses is a one-tailed test of significance appropriate?

A) Teaching method A and teaching method B will produce different student achievement gains.
B) There is no difference in the effects of Head Start programs on the achievement gains of low-income Caucasian and African-American children.
C) The mean IQ score of a sample is significantly different from the population mean.
D) Children receiving a dietary supplement will show greater achievement gains than those receiving a placebo.

Question 22

Use of a one-tailed versus a two-tailed t test to test the significance of the difference between two sample means depends on whether

A) there is expected overlap between the error curves of the two sample distributions.
B) the difference is expected to be in one direction only.
C) the size of the samples is proportional to their respective population sizes.
D) research participants were matched or chosen randomly.

Question 23

If any of the assumptions underlying the t test are violated to a modest extent,

A) the t test is still likely to yield an accurate estimate of statistical significance.
B) the t test will yield an accurate estimate of statistical significance, but only if the test is two-tailed.
C) a nonparametric test of statistical significance must be used instead.
D) no test of statistical significance should be performed.

Question 24

The t test for correlated means should be used when

A) the group possessing the characteristic of interest and the comparison group are randomly selected.
B) subgroups are to be formed within the group possessing the characteristic of interest.
C) the group possessing the characteristic of interest and the comparison group are matched on another variable.
D) researchers wish to determine the magnitude of the relationship between independent and dependent variables.

Question 25

In comparing the mean performance of 25 delinquents and 25 matched nondelinquents, one should use the

A) t test for independent means.
B) z distribution.
C) t test for a single mean.
D) t test for correlated means.

Question 26

Match each condition below with the test most appropriate under that condition To determine whether:

-the sample mean differs significantly from a specified population mean.

A) t test for correlated means
B) t test for a single mean
C) chi-square test
D) analysis of variance
E) t test for independent means

Question 27

Match each condition below with the test most appropriate under that condition To determine whether:

-the mean achievement of a random sample of delinquents differs significantly from that of a randomly selected control group.

A) t test for correlated means
B) t test for a single mean
C) chi-square test
D) analysis of variance
E) t test for independent means

Question 28

Match each condition below with the test most appropriate under that condition To determine whether:

-the mean anxiety scores of high, middle, and low achievers are significantly different.

A) t test for correlated means
B) t test for a single mean
C) chi-square test
D) analysis of variance
E) t test for independent means

Question 29

Match each condition below with the test most appropriate under that condition To determine whether:

-high, middle, and low achievers have different high school dropout rates.

A) t test for correlated means
B) t test for a single mean
C) chi-square test
D) analysis of variance
E) t test for independent means

Question 30

Match each condition below with the test most appropriate under that condition To determine whether:

-the mean achievement of two groups of students matched on aptitude differs significantly.

A) t test for correlated means
B) t test for a single mean
C) chi-square test
D) analysis of variance
E) t test for independent means

Question 31

Use of the t test for a single mean requires

A) a normal score distribution.
B) homogeneity of variance.
C) random selection of research participants.
D) knowledge of the population mean.

Question 32

If a researcher wishes to determine whether a sample mean differs from the population mean, the appropriate statistic is the

A) chi-square.
B) t test for independent means.
C) t test for a single mean.
D) analysis of covariance.

Question 33

The F ratio is computed to determine

A) whether the members of a sample have been randomly selected from a population.
B) the statistical significance of the difference between the mean scores of two or more groups.
C) whether two or more groups have different frequency distributions on a categorical variable.
D) all of the above.

Question 34

Analysis of variance focuses on the

A) between-groups variance.
B) within-groups variance.
C) ratio of between-groups variance to within-groups variance.
D) mean scores of each group in the analysis.

Question 35

The purpose of the t test for multiple comparisons is to

A) adjust for statistically significant differences that occur by chance because the comparison groups are compared on many variables.
B) provide an alternative approach to multiple comparisons when the assumptions underlying analysis of variance are not satisfied.
C) enable the use of analysis of variance when the number of comparison groups exceeds five.
D) all of the above.

Question 36

Analysis of covariance is the preferred test rather than analysis of variance when

A) there are initial differences between the samples on a measured variable related to the dependent variable.
B) there are extreme differences in the sample variances.
C) the scores of the groups being compared are in the form of categorical frequency counts.
D) the scores of the groups being compared are in the form of ranks.

Question 37

In conducting a MANOVA on a cluster of five related variables, a researcher obtains an F value that is significant at the .01 level. What is the next step that should be taken in analyzing the data?

A) Reexamine the cluster to see if any variables should be removed.
B) Test for the equality of group dispersions.
C) Compute a mean vector score for each variable.
D) Conduct an ANOVA on each dependent variable in the cluster.

Question 38

The test for homogeneity of independent variances, the test for homogeneity of related variances, and theF maximum test for homogeneity of variance are all forms of

A) analysis of covariance.
B) nonparametric statistics.
C) analysis of variance.
D) multivariate analysis of variance.

Question 39

Compared to parametric statistics, nonparametric statistics

A) have less statistical power.
B) have more statistical power.
C) involve many more assumptions about score distribution in the population being studied.
D) cannot be used to determine whether a null hypothesis can be rejected.

Question 40

Chi-square should be used when data are in the form of

A) ranks.
B) standard scores.
C) continuous scores.
D) category frequencies.

Question 41

The Mann-Whitney U test, the Wilcoxon signed rank test, and the Kruskal-Wallis test

A) all are forms of chi-square tests.
B) are preferred when the research data are rank or interval scores that grossly violate parametric assumptions of normal distribution and homogeneity of variance.
C) are used to compare the score distributions of three or more groups.
D) all of the above.

Question 42

Briefly describe the different role of description in descriptive research and in causal-comparative research.

Question 43

Name two descriptive statistics that often are calculated in a descriptive research study.

Question 44

A researcher makes direct observations of the classroom question-asking behavior of third graders. What type of descriptive research does this study exemplify?

Question 45

In longitudinal research, describe how samples are selected for (a) cohort, (b) trend and (c) panel studies, and provide an example of each sampling procedure.

Question 46

Lewis Terman's study of giftedness involved repeated measurement of a panel of children identified as geniuses at various stages in their lives. What is one weakness and one strength of this study design?

Question 47

Describe the major problem with the sampling procedure used in a cross-sectional research design, and provide an example that illustrates that problem.

Question 48

A researcher finds a correlation between length of teaching experience (variable X) and the principal's ratings of teaching effectiveness (variable Y) for the teachers employed in the principal's school. Give two plausible interpretations of this finding.

Question 49

Describe two approaches to selecting a suitable comparison group for a causal-comparative research study.

Question 50

List three assumptions about the obtained scores in a causal-comparative study that should be satisfied in order to use the t test.

Question 51

Under what condition should the researcher use the t test for correlated means in a causal-comparative study, and what effect will use of this type of t test have on the likelihood of finding a significant difference between groups on the variable of interest?

Question 52

A researcher wishes to determine whether the sample's mean score on a standardized achievement test is different from the mean score of the norming group. What type of t test is most appropriate in this situation?

Question 53

A researcher conducts a study in which 90 high school sophomores are randomly assigned to three algebra classes. Class A is taught by a lecture method, class B by a cooperative learning method, and class C by an individualized method. Algebra achievement, the dependent variable, is measured at the end of the course. What statistical significance test is most appropriate for comparing the mean algebra achievement scores of the three groups? If this test yields a value that is significant at the .05 level, how would you interpret this finding?

Question 54

What is the major difference between analysis of variance (ANOVA) and analysis of covariance (ANCOVA)?

Question 55

Three assumptions about population parameters should be checked when deciding to use parametric statistics to analyze causal-comparative data rather than nonparametric statistics. List one of these assumptions.

Question 56

You have hypothesized that dogmatism is distributed among army recruits so as to approximate a normal distribution (or normal probability curve). You randomly select 1,000 recruits at a basic training center and administer the Rokeach Dogmatism Scale. Scores of the 1,000 recruits are sorted into five categories ranging from high dogmatism to low dogmatism. What statistical procedure would be most appropriate to test your hypothesis, and why?

Question 57

Describe a suitable comparison group for each of the following defined groups in a causal-comparative study.
a. The discipline styles used by parents of high school students with high scores on an authoritarianism scale
b. The extent to which female college students who are in leadership roles in campus organizations report childhood experiences in which adults encouraged them to display leadership
c. The educational aspirations of fourth-grade Pueblo Indian boys who have attended school only on the reservation

Question 58

Two researchers are interested in doing a study to provide descriptive information to a school district about homelessness among students attending the district's schools. One researcher is a quantitative researcher, and the other is a qualitative researcher. Assume that they are able to obtain information from school records about all the students who are categorized as homeless. Describe (a) several types of information that the quantitative researcher would be more likely to report, and (b) several types of information that the qualitative researcher would be more likely to report.

Question 59

A researcher studies the difference in reading achievement of third-grade children who learned to read by the phonics method and those who learned to read by the whole language method of teaching reading. Children in the two groups are matched on an aptitude measure that has been found to correlate with measures of reading achievement. What type of t test should the investigator use to analyze the results, and why? If the t value is significant, what interpretation can the researcher make?

Question 60

You hypothesize that low-achieving children who receive special tutoring will make significantly greater reading achievement gains than those who spend an equal amount of time engaged in solitary reading in the regular classroom. Would this hypothesis be tested with a one-tailed or a two-tailed test of significance, and why?

Question 61

You hypothesize that low-achieving children who receive special tutoring in reading skills will make significantly greater gains than low-achieving children who do not receive the special tutoring on a measure of reading achievement and also on measures of three types of academic achievement (mathematics, social studies, and physical science). What test of statistical significance would you use in this situation?

Question 62

A researcher is interested in learning whether attendance at a high school with strict rules of discipline and conduct tends to produce authoritarian students. He locates two high schools that are located in very similar neighborhoods serving similar racial and socioeconomic groups. School A promotes strict rules of discipline and conduct, and most of its teachers and administrators have higher than average scores on a measure of authoritarianism. School B employs flexible rules of discipline and conduct, has a high level of student involvement in formulating school rules, and primarily employs teachers and administrators with lower than average scores on a test of authoritarianism.
a. In a pilot study the researcher randomly selects 25 students from each school and administers a 40-item scale of authoritarianism. What statistical technique should be used to analyze these data?
b. Each high school has 12 extracurricular programs that are classified by a group of independent judges as either high, average, or low in terms of authoritarian structure and activities. The researcher wishes to compare the proportion of students in each school who participated in each of the three types of programs. What statistical technique should be used to make this comparison?
c. In checking school records, the researcher found that most seniors in the two high schools had been administered the California F scale (a measure of authoritarianism) during their last year of junior high school. He decides to administer a different but similar authoritarianism measure to seniors who had taken the earlier test, and then see how the senior-level authoritarianism scores differed between students in the two schools when their earlier scores were taken into consideration. What statistical technique should be used?