Deck 5: Statistical Techniques

Question 1

Qualitative researchers differ from quantitative researchers in that they

A) collect quantitative data, but do not analyze it statistically.
B) do not use descriptive statistics to represent what is happening in a situation.
C) use statistical analysis as a supplement to interpretive analysis.
D) avoid collection of any quantitative data.

Answer 1

use statistical analysis as a supplement to interpretive analysis.

Question 2

Match each type of score named below with the appropriate example.

-High-anxiety, moderate anxiety, and low-anxiety students

A) Derived score
B) Categories
C) Rank score
D) Continuous score
E) Standard score

Answer 2

Categories

Question 3

Match each type of score named below with the appropriate example.

-High school graduation standing

A) Derived score
B) Categories
C) Rank score
D) Continuous score
E) Standard score

Answer 3

Rank score

Question 4

Match each type of score named below with the appropriate example.

-Verbal IQ = 120

A) Derived score
B) Categories
C) Rank score
D) Continuous score
E) Standard score

Question 5

Match each type of score named below with the appropriate example.

-Grade equivalent for scores on a reading test

A) Derived score
B) Categories
C) Rank score
D) Continuous score
E) Standard score

Question 6

Match each type of score named below with the appropriate example.

-Stanine scores for a reading test

A) Derived score
B) Categories
C) Rank score
D) Continuous score
E) Standard score

Question 7

The assignment of a score of 1 to male students and a score of 2 to female students is an example of a(n)

A) nominal scale.
B) ordinal scale.
C) interval scale.
D) ratio scale.

Question 8

Age equivalents, grade equivalents, and normal curve equivalent scores are all

A) test statistics.
B) standard scores.
C) derived scores.
D) all of the above.

Question 9

The difference between continuous scores and discrete scores is that

A) continuous scores are derived from interval scales, whereas discrete scores are derived from ratio scales.
B) continuous scores are derived from ratio scales, whereas discrete scores are derived from interval scales.
C) continuous scores, but not discrete scores, can be converted to percentiles.
D) continuous scores, but not discrete scores, can have fractional values.

Question 10

Match each statistic named below with the type of descriptive statistic under which it is classified.

-Standard deviation

A) Measures of central tendency
B) Measures of variability
C) Correlational statistics

Question 11

Match each statistic named below with the type of descriptive statistic under which it is classified.

-Bivariate coefficient

A) Measures of central tendency
B) Measures of variability
C) Correlational statistics

Question 12

Match each statistic named below with the type of descriptive statistic under which it is classified.

-Median

A) Measures of central tendency
B) Measures of variability
C) Correlational statistics

Question 13

Match each statistic named below with the type of descriptive statistic under which it is classified.

-Variance

A) Measures of central tendency
B) Measures of variability
C) Correlational statistics

Question 14

Correlational statistics serve the purpose of determining

A) the degree of relationship between two or more variables.
B) the degree to which the distribution of scores on a variable follow the normal probability curve.
C) whether the mean or median is the more appropriate descriptive statistic.
D) the magnitude of the difference in the standard deviations of two or more variables.

Question 15

The median of a group of scores

A) is a measure of variability.
B) reflects average performance less accurately than the mean when the distribution of scores is skewed.
C) is the middle score in the distribution of scores.
D) is an inferential statistic.

Question 16

Compared to the median, the mean

A) is less stable across samples drawn from the same population.
B) is more useful in analyses involving inferential statistics.
C) provides a more accurate representation of average performance.
D) is easier to compute.

Question 17

The mean and the, taken together, give a good description of the nature of the group beingstudied.

A) percentile distribution
B) median
C) standard deviation
D) score range

Question 18

If a score distribution forms a normal curve,

A) each standard deviation above or below the mean will include a fixed percentage of scores.
B) the standard deviation can be calculated from the mean alone.
C) the median is a better measure of central tendency than the mean.
D) the mode is a better measure of central tendency than the mean.

Question 19

Skewness means that

A) the distribution of scores on a measure is not symmetrical.
B) the mean will be in the direction of the more extreme scores in the score distribution.
C) the median and mean will differ.
D) all of the above.

Question 20

Compared with the use of a p value of .05, use of a p value of .01

A) provides a more stringent test of the research hypothesis.
B) makes a Type 1 error less likely.
C) increases the likelihood of accepting the null hypothesis when there is a real population difference.
D) all of the above.

Question 21

A test of the statistical significance of the difference between two sample means results in a t value that is significant at the .01 level. Which of the following is not a correct interpretation of this finding?

A) The probability is 99 percent that the samples were not drawn from the same population.
B) The mean difference exceeds the mean difference we would find once in 100 samples if the population mean difference was zero.
C) The null hypothesis can be rejected at the .01 alpha level.
D) It is probable that the difference between the sample means reflects a true difference between the population means.

Question 22

When we plan a research project, we decide on the level of statistical significance we will use to reject the null hypothesis. This level is called the

A) probability value.
B) t level.
C) alpha level.
D) Type I value.

Question 23

One way to increase statistical power is to

A) select a p value of .01 instead of .05.
B) select a larger sample.
C) decrease the risk of making a Type 1 error.
D) use a two-tailed test of statistical significance rather than a one-tailed test.

Question 24

Statistical power is

A) the power of a particular statistical test to detect a significant difference or other effect.
B) the ratio between the significance level and the effect size.
C) equivalent to the number of research participants included in the study.
D) the probability that a particular statistical test will lead to rejection of a false null hypothesis.

Question 25

The only factor that a researcher cannot manipulate in order to increase statistical power is

A) effect size.
B) sample size.
C) the p value for rejecting the null hypothesis.
D) the specification of directionality in the research hypothesis.

Question 26

The mean and standard deviation of a sample's distribution of scores on a measure can be used to estimate

A) a range of values that is likely to include the true population mean.
B) the probability that the sample and population means truly differ.
C) the probable margin of error in an individual's obtained score on a measure used in the study.
D) statistical power.

Question 27

The fifth edition of the Publication Manual of the American Psychological Association states that the best statistics to use in reporting the results of quantitative analyses are

A) tests of statistical significance.
B) confidence intervals.
C) effect sizes.
D) means and standard deviations..

Question 28

The replication of a finding from one study by a subsequent study is generally considered to be

A) more supportive of a knowledge claim than a finding from one study that has strong statistical power.
B) less supportive of a knowledge claim than a finding from one study that has strong statistical power.
C) unnecessary because of the development of confidence-limit methodology.
D) unnecessary because of the development of effect-size methodology.

Question 29

A comparison of the mean scores of two groups on a test yields an effect size (ES) of 1.41. An ES of this magnitude means that

A) the difference between the mean scores of the two groups is trivial.
B) the difference between the mean scores of the two groups is large.
C) the difference between the standard deviations of the two groups is trivial.
D) both a and c above are correct.

Question 30

The practical significance of a research finding usually is best determined by

A) tests of statistical significance.
B) tests of statistical power.
C) confidence intervals.
D) effect sizes.

Question 31

One of the advantages of stem-and-leaf displays of data is that they facilitate the detection of outliers. What is an outlier?

A) An individual whose score differs markedly from the other scores in the display.
B) An individual whose score differs substantially from the other scores on any of the leaves of the display.
C) An individual whose score is either at the top or at the bottom of the display's stem.
D) An individual who is outside of the target population.

Question 32

Stem-and-leaf displays are useful because they make it easy to

A) see the distribution of scores on a measure.
B) detect outliers.
C) see unexpected patterns in the data.
D) all of the above.

Question 33

A diagram that shows the relationship between scores on two measures of continuous variables is called a

A) bar graph.
B) histogram.
C) stem and leaf display.
D) pictogram.

Question 34

The difference between histograms and bar graphs is that

A) only histograms show the relationship between two variables.
B) only bar graphs show the relationship between two variables.
C) only bar graphs include categorical scores.
D) only bar graphs can have bars that touch each other.

Question 35

The best approach for dealing with the problem of missing data is to

A) have the same person carry out all aspects of the data analysis.
B) increase the number of required data collection sessions.
C) eliminate incomplete cases from the statistical analysis.
D) take extra care beforehand to ensure that all required data are collected.

Question 36

Multilevel analysis refers to a procedure in which

A) students at different grade levels are compared.
B) the analysis of data is carried out at several levels, such as individual students, classrooms, and schools.
C) the statistical analysis begins with simple statistical tests, such as chi-square, and progresses to more sophisticated tests, such as ANOVA.
D) the appropriate probability level for statistical analysis is determined in advance.

Question 37

In statistical analyses made with a computer, a command file is

A) the raw data entered into the computer program.
B) a list of the statistical procedures from which one can choose in analyzing the data.
C) a summary of the results of the statistical analysis.
D) a list of the computer instructions that were used to analyze a data set.

Question 38

Why are quantitative researchers more likely than qualitative researchers to use statistics in their investigations?

Question 39

A researcher collects data from a sample of school principals on the following variables: (a) age; (b) gender; (c) whether the principal's school is at the elementary, middle, or high school level; and (d) their standingrelative to each other on a measure of effectiveness. Identify the type of score that would be yielded by a measure of each of these variables.

Question 40

If researchers want to know whether boys are more homogeneous in visual-spatial ability than girls, would they use a measure of central tendency, a measure of variability, or a correlational method?

Question 41

State one advantage and one disadvantage of the median as a measure of central tendency.

Question 42

If a sample's scores on a test form a normal curve, what is one piece of information that the standard deviation provides?

Question 43

What does a test of statistical significance of the difference between the mean scores of two groups tell researchers about their null hypothesis?

Question 44

(a) What is statistical power? (b). Briefly describe two factors that increase statistical power.

Question 45

A researcher compares the scores of 1,000 randomly selected urban eighth grade students with 1,000 randomly selected rural eighth grade students on a 30-item test of mathematical understanding. The urban sample obtains a mean raw score of 12.6, while the rural sample obtains a mean raw score of 13.2. The standard deviation is 5.1. The difference between the mean scores is significant at the .05 level. Would you consider this difference to be of practical significance? Why or why not?

Question 46

Explain what confidence limits reveal about the population values based on the findings for a research sample.

Question 47

What are two advantages of organizing data into a stem-and-leaf display for exploratory data analysis?

Question 48

Several students in a researcher's sample take a test administered at the start of the study but do not take the test administered at the end of the study. Give one reason why this creates a problem for the research procedure.

Question 49

A researcher administers a test of musical knowledge to all students in 30 classrooms and then compares the performance of the boys and girls on this test. If the unit of statistical analysis is the classroom, how would the comparison be made?

Question 50

State one reason why it is desirable to retain one's raw research data for a period of time after a study has been completed.

Question 51

You want to know the average gain in reading speed that is made by students who take a speed-reading course. A reading test was given to 250 students before and after they took the course. In looking over the scores, you notice that most students gained 100 to 300 words per minute. However, there are about 20 students who made outstanding gains, ranging from 600 to 1,000 words per minute. Would the mean or the median be a more accurate measure of central tendency, and why?

Question 52

A researcher finds that the mean score of scholarship students on a final exam in a university history course is 89.7, and the mean score of other students is 85.3. A t test of this difference yields a p value of .01. What does the p value tell us about the null hypothesis? Also, what generalization does the p value allow us to make?

Question 53

A researcher has completed a study in which an experimental group of 60 college students was formed into pairs and was requested to ask each other questions in preparation for an examination. A control group of 60 students was requested to quiz themselves in preparation for the same examination. Students' scores on the examination formed the dependent variable. What is the appropriate unit of statistical analysis for determining whether the examination performance of the experimental group differed from that of the control group?