Deck 13: Describing Data

Question 1

The search for patterns in your data is called:

A) inferential statistical analysis.
B) descriptive data analysis.
C) exploratory data analysis.
D) exploratory pattern making.

Answer 1

exploratory data analysis.

Question 2

Why is it a good idea to explore your data using EDA techniques before you conduct any statistical tests?

A) They help spot serious defects in your data that may warrant taking corrective action before you proceed to the inferential analysis.
B) They can help you determine which summary statistics would be appropriate for a given set of data.
C) They may reveal unsuspected influences.
D) All of the answers are correct.

Answer 2

All of the answers are correct.

Question 3

If categories are represented by numbers (e.g., 1 = Democrat, 2 = Republican), the categories are said to be:

A) dummy-coded.
B) nominalized.
C) pseudonumbered.
D) transformed.

Answer 3

dummy-coded.

Question 4

Identify a true statement about an unstacked format, which is a way of organizing data from experimental or quasi-experimental designs.

A) This format is appropriate when data include multiple independent or dependent variables.
B) This format fails to display treatment summary statistics in a simple way.
C) In this format, multiple columns of subject numbers are needed.
D) In this format, a separate column is created for the scores from each treatment.

Question 5

Which of the following is a disadvantage of a stacked format?

A) It makes it difficult to include additional columns to indicate the observed values of additional variables.
B) It fails to accommodate complex designs involving several quasi-independent variables.
C) It does not provide a simple way to display treatment summary statistics.
D) It is not an acceptable format for many computer statistical analysis packages.

Question 6

Dummy coding involves:

A) assignment of numbers to the levels of a qualitative independent variable.
B) creating new variables based on the values of the old ones.
C) categorizing dependent variables according to their magnitudes.
D) None of the answers is correct.

Question 7

For quantitative data (e.g., the number of milligrams of a drug), coding your data involves:

A) dummy coding.
B) transferring each subject's score to a computer coding sheet.
C) assigning numeric values to categorical data.
D) None of the answers is correct.

Question 8

When coding dependent variables, you should not worry about creating new variables or with making special categories because:

A) computers cannot read such composite variables.
B) they are rarely used in data analysis.
C) most statistical analysis software have commands that lets one manipulate data in a variety of ways.
D) All of the answers are correct.

Question 9

After entering data, _____.

A) you need not check for data errors because the computer can do it for you
B) the computer recognizes errors and corrects the data automatically
C) you need not check for data errors because they are rare
D) you should carefully check your data file for errors because the computer cannot detect incorrectly entered data

Question 10

A limitation of using grouped data is that:

A) the average score may not represent the performance of individual subjects in a group.
B) a curve resulting from plotting averaged data may not reflect the true nature of the psychological phenomenon being studied.
C) grouping scores adds error variance to the data.
D) the average score may not represent the performance of individual subjects in a group, and a curve resulting from plotting averaged data may not reflect the true nature of the psychological phenomenon being studied.

Question 11

Examining individual data shows how each subject performed in a study when:

A) you have repeated measures of the same behavior.
B) several subjects per treatment group provide data measured on an interval scale.
C) more than five subjects are included in each treatment group.
D) data are measured on a ratio scale.

Question 12

When organizing data, a good strategy to adopt is:

A) restrict yourself to looking at grouped data.
B) restrict yourself to looking at individual data.
C) to look at both grouped and individual data.
D) to look at neither grouped nor individual data.

Question 13

The horizontal axis of a graph is called the _____.

A) abscissa or x-axis
B) abscissa or y-axis
C) ordinate or x-axis
D) ordinate or y-axis

Question 14

When graphing data from an experiment, levels of an independent variable are normally represented along the:

A) y-axis.
B) x-axis.
C) z-axis.
D) ordinate.

Question 15

In a bar graph, the length of each bar reflects the value of the:

A) independent variable.
B) abscissa.
C) dependent variable.
D) quasi-independent variable.

Question 16

A bar graph is the best method of graphing data when your independent variable is:

A) scaled on an interval scale.
B) scaled on a ratio scale.
C) continuous.
D) categorical.

Question 17

Which of the following statements is true about a bar graph?

A) In this graph, each pair of scores is represented as a point on the graph.
B) It is most appropriate when the independent variable is continuous and quantitative.
C) In this graph, the width of each bar reflects the value of the dependent variable.
D) It presents data as bars extending away from the axis representing the independent variable.

Question 18

_____ are more appropriate than bar graphs when your independent variable is continuous and quantitative.

A) Line graphs
B) Scatter plots
C) Pie graphs
D) All of the answers are correct.

Question 19

Line graphs are appropriate when one wants to illustrate:

A) the main effects of variables.
B) categorical relationships among variables.
C) dichotomized data splits.
D) functional relationships between variables.

Question 20

A(n) _____ curve is relatively flat at first and becomes progressively steep as it moves along the x-axis.

A) negatively accelerated
B) non-accelerated
C) positively accelerated
D) increasing

Question 21

A curve that is steep at first but becomes progressively flatter as it moves along the x-axis is:

A) negatively accelerated.
B) positively accelerated.
C) nonmonotonic.
D) non-accelerated.

Question 22

When a curve levels off at some maximum or minimum value, the function is said to be _____ at that value.

A) nonlinear
B) nonmonotonic
C) positively accelerated
D) asymptotic

Question 23

A curve that represents a uniformly increasing or decreasing function is said to be:

A) monotonic.
B) nonmonotonic.
C) cubic.
D) asymptotic.

Question 24

Pairs of scores from a correlational study are usually represented as points on a:

A) histogram.
B) scatter plot.
C) line graph.
D) pie graph.

Question 25

If your data are in the form of proportions or percentages, then a good type of graph to represent the value of each category in an analysis would be a:

A) line graph.
B) histogram.
C) pie graph.
D) bar graph.

Question 26

Graphing data, rather than presenting them in a table, is important when you:

A) want to show relationships clearly.
B) are choosing appropriate statistics.
C) are contemplating not using descriptive statistics.
D) want to show relationships clearly and when choosing appropriate statistics.

Question 27

A set of mutually exclusive categories (classes) together with a count of the number of data values falling into each category is termed a:

A) sorted list.
B) cumulative distribution.
C) frequency distribution.
D) scatter plot.

Question 28

_____ resemble bar graphs, with each bar representing a class, and a given bar's length indicating the frequency of scores falling within its range.

A) Histograms
B) Scatter plots
C) Exploded pie graphs
D) Stemplots

Question 29

In a stemplot of scores ranging from 11 to 83, a score of 42 would be located at a stem value of _____.

A) 4
B) 6
C) 40
D) 2

Question 30

Stemplots have the advantage over histograms of:

A) plotting the shape of a distribution.
B) preserving all the actual values present in the data.
C) determining the center of a distribution.
D) determining the spread of a distribution.

Question 31

When examining a histogram or stemplot of your data, you should:

A) locate the center of the distribution along the scale of measurement.
B) note the spread of the scores.
C) note the overall shape of the distribution and look for any gaps or outliers.
D) All of the answers are correct.

Question 32

Extreme scores that lie far from the others in a distribution are called:

A) deviants.
B) distant scores.
C) outlaws.
D) outliers.

Question 33

A distribution is _____ if a long tail goes off to the right, upscale.

A) normal
B) positively skewed
C) negatively skewed
D) bimodal

Question 34

A distribution contains the following scores: 1, 1, 2, 2, 2, 3, 3, 4, 4, 5. Its mode is:

A) 5.
B) 3.
C) 2.
D) 4.

Question 35

Identify a true statement about mode.

A) It is the most frequent score in a distribution.
B) It is the middle score in a distribution.
C) It is the most widely used measure of center.
D) It is the simplest and least informative measure of spread.

Question 36

Although mode is simple to calculate, it is limited because it:

A) is insensitive to extreme scores.
B) is difficult to compute.
C) is inappropriate for use with interval data.
D) does not take into account the values of scores outside of the most frequent score.

Question 37

A distribution contains the following scores: 1, 1, 2, 2, 2, 3, 3, 4, 4, 5. Its mean is:

A) 2.5.
B) 2.7.
C) 2.0.
D) 4.0.

Question 38

In a distribution with an even number of scores, the median is determined by:

A) finding the most frequent score in the top half of the distribution and averaging it with the most frequent score in the bottom half of the distribution.
B) finding the arithmetic average of the entire distribution and dividing it by two.
C) averaging the middle pair of scores.
D) finding the most frequent score.

Question 39

The median is a rather insensitive measure of center because it:

A) is difficult to calculate.
B) does not take into account the magnitudes of the scores above and below it.
C) cannot be used with interval data.
D) All of the answers are correct.

Question 40

The _____ is the most sensitive measure of center because it takes into account all scores in a distribution when it is calculated.

A) median
B) mode
C) interquartile range
D) arithmetic average

Question 41

Which of the following is the major advantage of the mean?

A) Its value is directly affected by the magnitude of each score in a distribution.
B) Its insusceptibility to the influence of outliers makes it highly reliable.
C) It is an appropriate measure of center when data are measured on an ordinal scale.
D) It is the most preferred measure whenever a distribution is strongly skewed.

Question 42

The mean is derived by:

A) finding the most frequent score in a distribution.
B) finding the middle score in an ordered distribution.
C) summing the scores in a distribution and dividing the sum by the total number of scores.
D) averaging the middle pair of scores in an ordered distribution.

Question 43

For data measured on a nominal scale, you are limited to using the _____ as your measure of center.

A) mean
B) mode
C) median
D) All of the answers are correct.

Question 44

If data are scaled on an interval or ratio scale, the mean becomes a less representative measure of center when:

A) there are more than 10 scores in a distribution.
B) there are less than 5 scores in a distribution.
C) the mean and median are equal.
D) the distribution of scores is strongly skewed.

Question 45

In a positively skewed distribution, the mean:

A) underestimates the center.
B) overestimates the center.
C) is as accurate a measure of the center as is the median.
D) accurately represents central tendency.

Question 46

The simplest and least informative measure of spread is the:

A) standard deviation.
B) variance.
C) range.
D) interquartile range.

Question 47

In the context of the measures of spread, the _____ is the average squared deviation from the mean.

A) standard deviation
B) variance
C) range
D) interquartile range

Question 48

The most popular measure of spread is the:

A) standard deviation.
B) variance.
C) interquartile range.
D) range.

Question 49

Which of the following measures of spread is easy to calculate and is resistant to the effects of skew and outliers?

A) The standard deviation
B) The variance
C) The interquartile range
D) The range

Question 50

Included in the five-number summary are the:

A) minimum, the first quartile, the median, the third quartile, and the maximum.
B) mean, the median, the mode, the standard deviation, and the interquartile range.
C) minimum, the interquartile range, the standard deviation, the range, and the maximum.
D) mean, the median, the interquartile range, the standard deviation, and the range.

Question 51

You can display the five-number summary graphically as a:

A) histogram.
B) bar graph.
C) boxplot.
D) scatter plot.

Question 52

The _____ of a correlation coefficient tells you the direction of a relationship, whereas the _____ tells you the degree of linear relationship between two variables.

A) magnitude; sign
B) value; magnitude
C) sign; magnitude
D) None of the answers is correct.

Question 53

The presence of outliers can affect the _____ of the Pearson r.

A) sign
B) magnitude
C) magnitude, sign, or both
D) None of the answers is correct.

Question 54

Which of the following is the most widely used measure of association and is appropriate when the dependent measures are scaled on an interval or a ratio scale?

A) The point-biserial correlation
B) The phi coefficient
C) The Spearman rank-order correlation
D) The Pearson r

Question 55

The measure of correlation to use when one variable is measured on an interval scale and the other is measured on a nominal scale is the:

A) Spearman rank-order correlation.
B) point-biserial correlation.
C) phi coefficient.
D) part correlation.

Question 56

The Spearman rank-order correlation is used when:

A) dependent variables are scaled on a ratio scale.
B) one variable being measured is on an interval scale and the other being measured is on a nominal scale.
C) both of the variables being correlated are measured on a dichotomous scale.
D) one wants to determine whether the relationship between variables is monotonic.

Question 57

With _____, you can estimate values of a variable based on knowledge of the values of others.

A) linear regression
B) the Pearson r
C) the phi coefficient
D) the coefficient of determination

Question 58

The best-fitting straight line on a scatter plot that minimizes the sum of the squared deviations of each data point from the line is called the:

A) standard error line.
B) optimal line.
C) least-squares regression line.
D) None of the answers is correct.

Question 59

In the formula that describes the regression line mathematically, b is:

A) a constant.
B) the regression weight.
C) a predicted score.
D) the same as the Pearson r.

Question 60

The difference between Y and $\hat{Y}$ is a:

A) residual.
B) remainder.
C) regression deviation.
D) difference.

Question 61

The _____ provides an estimate of the amount of error in prediction.

A) standard deviation
B) variance
C) regression weight
D) standard error of estimate

Question 62

The standard error of estimate increases as:

A) beta decreases.
B) the strength of the relationship between X and Y increases.
C) the strength of the relationship between X and Y decreases.
D) the constant in the regression equation increases.

Question 63

By squaring the correlation coefficient, you can:

A) determine how many deviant scores are there in your distributions.
B) obtain the strength of the causal relationship between variables.
C) obtain an index of the amount of variation in one variable that can be accounted for by variation in the other.
D) determine how much unit change in X you can expect with a unit change in Y.

Question 64

A high value of the coefficient of nondetermination means that:

A) there is a causal relationship between variables.
B) performing linear regression is unnecessary.
C) you can expect a large change in X as a function of Y.
D) None of the answers is correct.

Question 65

A correlation matrix results when you:

A) compute all possible correlations among a number of variables.
B) enter your scores into a linear regression analysis.
C) compute beta weights for all your variables.
D) None of the answers is correct.

Question 66

Organization of data depends on the research design, the number of variables that were observed and recorded, and how observations were grouped or subdivided.

Question 67

Grouped data may not reflect the true nature of the psychological phenomenon being studied.

Question 68

It is a good idea to look at both grouped and individual data where possible.

Question 69

A functional relationship is one in which the value of a dependent variable varies as a function of the value of an independent variable.

Question 70

A monotonic relationship is one in which a functional relationship changes direction several times.

Question 71

Graphing data on a scatter plot is helpful when one intends to calculate a measure of correlation.

Question 72

Proportions and percentages can best be represented on a pie graph.

Question 73

When displaying frequency distributions, stemplots are especially useful for large data sets.

Question 74

A measure of center, also known as a measure of central tendency, gives a single score that represents the general magnitude of scores in a distribution.

Question 75

In a negatively skewed distribution, the mean overestimates the center.

Question 76

In the context of the Pearson r, a negative correlation indicates an inverse relationship between two sets of scores.

Question 77

The Pearson r is insensitive to the shapes of the score distributions.

Question 78

The phi coefficient ( $\phi$ ) is used when both of the variables being correlated are measured on a dichotomous scale.

Question 79

$\hat{Y} = a + bX$ is the equation for the least-squares regression line.

Question 80

The standard error of estimate provides a measure of the amount of variance that two variables being tested share.