Ch. 16: Data Analysis: Descriptive and Inferential Statistics

The nurse researcher is interested in estimating how reliably data can be used to generalize the findings of a study. In order for this to occur, the researcher should use what? a. Inferential statistics b. Descriptive statistics c. Nonparametric statistics d. Mathematical statistics

ANS: A Feedback A Inferential statistics are used by researchers to make predictions and generalize findings on the basis of data. B Descriptive statistics are used to describe and summarize data by reducing data to manageable proportions by summarizing them. C Nonparametric statistics are a type of inferential statistics. D The term mathematical statistics is a circular answer.

The nurse researcher is using a research study that involved categories of patient weights. The level of measurement that would be most appropriate in this study is what? a. Nominal b. Ordinal c. Interval d. Ratio

ANS: A Feedback A The data are represented by nominal-level data, categories, or classifications. B The data are represented by nominal-level data, categories, or classifications. C The data are represented by nominal-level data, categories, or classifications. D The data are represented by nominal-level data, categories, or classifications.

The nurse researcher's data contain extremes of high and low scores. The measure of central tendency that should be used in order to be least affected by extremes of high and low scores is what? a. Mean b. Mode c. Median d. Percentile

ANS: A Feedback A The mean is affected by every score and the least affected by chance. The larger the sample size, the less affected the mean will be by a single score. The mean is the most stable, least affected by extremes, and most useful for other calculations. B The mode is unstable and can fluctuate widely from sample to sample from the same population. C The median is not sensitive to extremes in high and low scores; it is less stable than the mean. D Percentile is not a measure of central tendency; the median is the 50th percentile.

The nurse researcher is using a nursing degree as a variable in a research study. The level of measurement most appropriate in this case is what? a. Nominal b. Ordinal c. Interval d. Ratio

ANS: A Feedback A This variable (type of nursing degree) is scaled using categorical data, nominal level. The type fits only one characteristic. B This variable (type of nursing degree) is scaled using categorical data, nominal level. The type fits only one characteristic. C This variable (type of nursing degree) is scaled using categorical data, nominal level. The type fits only one characteristic. D This variable (type of nursing degree) is scaled using categorical data, nominal level. The type fits only one characteristic.

Which data measurement scales typically use nonparametric tests of significance? (Select all that apply.) a. Nominal b. Ordinal c. Interval d. Ratio

ANS: A, B Feedback Correct Nonparametric tests of significance are used when data are measured on a nominal or ordinal scale. Nonparametric tests of significance are used when data are measured on a nominal or ordinal scale. Incorrect Parametric statistics are used when data are scaled at the interval or ratio level scale. Parametric statistics are used when data are scaled at the interval or ratio level scale.

Which measures of central tendency allow for relative rankings? (Select all that apply.) a. Mode b. Range c. Median d. Mean e. Semiquartile range f. Percentile g. SD

ANS: A, C Feedback Correct The mode will allow for relative rankings. The median will allow for relative rankings. Incorrect This is an ordinal measure of variability. The mean will not allow for relative rankings. This is an ordinal measure of variability. This is an ordinal measure of variability. This is an ordinal measure of variability.

Which measures of central tendency allow for rank ordering with equal intervals? (Select all that apply.) a. Mode b. Range c. Median d. Mean e. Semiquartile range f. Percentile g. SD

ANS: A, C, D Feedback Correct The mode will allow for rank ordering with equal intervals. The mode will allow for rank ordering with equal intervals. The mode will allow for rank ordering with equal intervals. Incorrect This is an ordinal measure of variability. This is an ordinal measure of variability. This is an ordinal measure of variability. This is an ordinal measure of variability.

A nurse researcher determines the assignment of a number to an object for measurement by verifying what? a. The objects vary within a narrow range to meet category requirements. b. All objects assigned a specific measurement number are similar to all other objects assigned the same number. c. The objects assigned to one specific number are positively related to the dependent variable of the study. d. The objects assigned to one specific number are positively related to the independent variable of the study.

ANS: B Feedback A A narrow range is not necessary to meet category requirements; the range may be large. B The assignment of a number or a value to a category that represents the category must be consistently applied to similar objects assigned the same number-value label. C The numbers assigned to objects are consistently applied; this relates to independent and dependent variables. D The numbers assigned to objects are consistently applied; this relates to independent and dependent variables.

The nurse researcher is interested in describing and summarizing the characteristics of the sample used in the research study. In order for this to occur, the researcher should use what? a. Inferential statistics b. Descriptive statistics c. Nonparametric statistics d. Mathematical statistics

ANS: B Feedback A Inferential statistics are used by researchers to make predictions and generalize findings on the basis of data. B Descriptive statistics are used to describe and summarize data by reducing data to manageable proportions by summarizing them. C Nonparametric statistics are a type of inferential statistics. D The term mathematical statistics is a circular answer.

The nurse researcher opts to group frequency data using large interval widths in a research study. Which problem may be encountered by using this type of interval widths? a. Overlap can occur. b. Patterns can be obscured. c. Means cannot be determined. d. Some data may not fit into any category.

ANS: B Feedback A Overlap is eliminated by defining the size of the group or interval width so that no score falls into two groups. B Very large interval widths lead to loss of data information and may obscure patterns in the data. C Frequency intervals do not generate means. D Data or a score generated on a variable will fit into one frequency interval defined by the investigator.

The nurse researcher is using the percentage of nurses holding various nursing degrees as a variable in a research study. Which type of statistics is most appropriate in this case? a. Inferential statistics b. Descriptive statistics c. Nonparametric statistics d. Mathematical statistics

ANS: B Feedback A Percentage represents descriptive statistics. B Percentage represents descriptive statistics. C Nonparametric statistics are used when testing hypotheses on data generated at nominal and ordinal units. D This is a circular answer and does not apply.

Which term represents the most frequent score in a frequency distribution? a. Mean b. Mode c. Median d. Percentile

ANS: B Feedback A The mean as an average represents interval- and ratio-level data. B The category with the highest frequency is the mode and is used often with nominal-level data. C The median is the middle score or the score when 50% of the scores are above it and 50% of the scores are below it. D Percentile represents the percentage of cases a given score exceeds.

A nurse researcher would opt to use factor analysis in which case? a. To control statistically for a variable and measure differences among group means b. To take a large number of variables and group them into a smaller number of factors c. To test the relationship between two variables measured at interval or ratio level data d. To measure the relationship between one outcome or dependent variable and several dependent variables

ANS: B Feedback A This answer defines analysis of covariance. B This is a definition of factor analysis. C This defines a correlation coefficient. D This answer defines multiple regression.

The nurse researcher opts to use interval-level data. The researcher's primary concern will be what? a. Categories must be mutually exclusive. b. Intervals between data categories must be equal. c. The rank of zero (0) represents the absence of the measured variable. d. Variance between assigned categories is qualitative rather than quantitative.

ANS: B Feedback A This response is true for all levels of data: nominal, ordinal, interval, ratio. B This is the defining characteristic of interval measurement. C The zero does not represent the total absence of a variable in interval-level data. D This response does not describe interval-level data.

The nurse researcher opts to use a measure of variability. Which measure of variability does the researcher recognize as being most unstable? a. Mean b. Median c. Range d. Semiquartile range

ANS: C Feedback A Mean is a measure of central tendency. B Median is a measure of central tendency. C The range is very likely to be changed by a single extreme score. D The semiquartile range is more stable than the range; it is less likely to be changed by a single extreme score.

The nurse researcher is interested in using nominal level data. The object or event that could be used if nominal-level data are of primary interest is what? a. Test scores b. Education level c. Marital status d. Weight

ANS: C Feedback A Test scores are best represented by interval-level data. B Education level is best represented by ordinal data: increasing rankings of education level. C Marital status is most often represented by nominal-level data. D Weight is represented by interval or ratio-level data.

In a study of nurses' willingness to care for patients with a contagious disease, it was found that the greater the nurses' spirituality, the greater the willingness to provide care. This finding represents what? a. No correlation b. Perfect correlation c. Positive correlation d. Negative correlation

ANS: C Feedback A The example suggests a correlation, not the absence of a correlation. B There is no indication in the example of a 1.0 correlation coefficient. C Both variables are correlated in the same (+, +) direction. D The variables are not correlated in opposite (+, -) directions.

The nurse researcher opts to use ratio-level data. The researcher's primary concern will be what? a. Categories must be mutually exclusive. b. Intervals between data categories must be equal. c. The rank of zero (0) represents the absence of the measured variable. d. Variance between assigned categories is qualitative rather than quantitative.

ANS: C Feedback A This response is true for all levels of data: nominal, ordinal, interval, ratio. B This is true for intervals between numbers for interval- and ratio-level data. C This is characteristic of ratio measurement. D This response does not describe interval-level data.

The nurse researcher finds that the data include a score reported to be in the 75th percentile. The researcher interprets this to mean what? a. The score is at the median. b. The score has exceeded the mode. c. The score is exceeded by only 25% of the cases. d. The score represents the average of all scores measured at this time.

ANS: C Feedback A You cannot conclude this on the basis of the data provided. B You cannot conclude this on the basis of the data provided. C The score of 75% is exceeded by only 25% of the cases. D This is not a logical answer.

Which characteristic is an essential criterion for categorizing an object, event, or fact as nominal level data? (Select all that apply.) a. The degree of the characteristic must be positively related to the number assigned. b. Rank orders must be calculated for meaningful interpretation. c. Categories must be mutually exclusive. d. Category ranges cannot overlap.

ANS: C, D Feedback Correct This is the definition of nominal-level data; categories are mutually exclusive. Category ranges must not overlap; each score fits into only one category. Incorrect This answer does not fit categorizing an object, event, or fact as nominal level. Rank orders apply to ordinal-level data, not to nominal-level data.

Which statements regarding measures of central tendency are accurate? (Select all that apply.) a. This classification of levels of measurement is scientifically inaccurate. b. Having the mean, mode, and median exactly the same indicates the measures have been correctly applied. c. They have great flexibility with regard to the amount of mathematical manipulation possible. d. They vary from sample to sample, even when the same trait or characteristic is measured.

ANS: C, D Feedback Correct Various inferential statistical tests employ mean, median, or mode in their calculations. Measures of central tendency vary from sample to sample, even when the same trait is measured. Incorrect The classification of levels of measurement is well accepted by the scientific community. This is an inaccurate statement.

The nurse researcher is interested in having the greatest flexibility possible in choosing statistical procedures. The level of measurement used to achieve this is what? a. Nominal b. Ordinal c. Interval d. Ratio

ANS: D Feedback A Nominal-level data allow the least amount of mathematical manipulation. B Ordinal-level data are limited in the amount of mathematical manipulation possible. C Interval-level data allow for a great amount of data manipulation. D Ratio measurement is the highest level of measurement; all mathematical procedures can be performed on data from ratio scales.

The nurse researcher is using respiratory rate as a variable in a research study. The level of measurement most appropriate in this case is what? a. Nominal b. Ordinal c. Interval d. Ratio

ANS: D Feedback A Rate is represented by a number representing frequency per minute. B Rate is represented by a number representing frequency per minute. C Rate is represented by a number representing frequency per minute. D Respiratory rate can have an absolute zero—absence of respirations.

A nurse researcher would opt to use measures of central tendency in which case? a. When there is a need to describe how reliably the findings can be generalized b. When there is a need to describe how much dispersion is in the sample c. When there is a need to describe the accuracy of predictions made d. When there is a need to describe the average member of the sample

ANS: D Feedback A This answer describes a function of inferential statistics. B This answer describes measures of variability. C This answer is incorrect. D Measures of central tendency yield a single number that describes the middle of the group and summarizes the members of a sample.

The nurse researcher is interested in describing the assignment of numbers to objects or events. This is best achieved in which way? a. Through use of prediction b. Through use of dispersion c. Through use of variability d. Through use of measurement

ANS: D Feedback A This answer does not fit the definition provided. B This answer represents a measure of variability description. C This answer represents a measure of dispersion (variability) description. D This answer fits the definition provided.

The nurse researcher should report the mean, mode, and median of a variable in a research study in which case? a. When the distribution is uneven. b. When the distribution is wide. c. When the distribution is narrow. d. When the distribution is symmetrical.

ANS: D Feedback A This is an inaccurate statement. B This is an inaccurate statement. C This is an inaccurate statement. D If the distribution of a variable of a sample is symmetrical and unimodal, the mean, mode, and median should be reported.

The alpha level that is interpreted by the nurse researcher as a highly statistically significant result is what? a. p = 0.20 b. p = 0.02 c. p = 0.002 d. p = 0.0002

ANS: D Feedback A This result is above the conventional 0.05 and is not statistically significant. B Although this result is statistically significant, it is not highly statistically significant. C This is a highly statistically significant result, but not the highest. D This is the most highly statistically significant result.