STUDY QUESTIONS * 1. Identify the variables that are measured at the nominal level in Table 1. * 2. Is age measured at the nominal level? Provide a rationale for your answer. * 3. What is the mode for marital status in the total sample and in the depressed group? * 4. How many of the depressed persons were unemployed? Would you have expected this finding? Provide a rationale for your answer. * 5. How many of the depressed persons were married? Would you have expected this finding? Provide a rationale for your answer. * 6. No depression was reported in the 29 to 38 age group. What are the possible reasons for this? * 7. How many subjects and what percentage of the total sample were 49 years of age or older? Round your answer to the nearest tenth of a percent (%). * 8. Nominal data categories should be exhaustive. In looking at Ethnicity in Table 1, are these ethnic categories exhaustive? Provide a rationale for your answer. * 9. For which variables in this table can a median be calculated? What does this tell us about the level of measurement of the variables? * 10. Are parametric or nonparametric statistical analysis techniques used to analyze nominal level data? Provide a rationale for your answer.
ANSWERS TO STUDY QUESTIONS * 1. In Table 1, gender, ethnicity, marital status, and employment status were measured at the nominal level. * 2. No, the age categories can be rank ordered and are therefore an example of ordinal level of measurement and not nominal level of measurement. * 3. The mode for marital status is married with a frequency of 80 subjects, which was 67% of the total sample. The mode for the depressed group was also married with a frequency of 32, which was 73% of the depressed group. * 4. 13 (29.5%) of the depressed persons were unemployed. Yes, you would expect that persons who have a comorbid condition and are experiencing depression might have a higher unemployment rate. In addition, unemployment might also lead to or increase depression. The evidence-based information on assessment, diagnosis, and management of depression is available online at http://www.guideline.gov. * 5. 32 (73%) of the depressed persons were married. This is a higher number than might be expected. This finding is contradictory to other studies on depression, which have found that persons who are in an intact marriage are less depressed than people who are divorced or single (http://www.ahcpr.gov). However, these subjects have comorbid conditions (cancer, MI, or stroke) or are caregivers, which might explain the number of married subjects with depression. * 6. One possible reason is that there are only 3 persons in that group and the number of subjects is too small to determine the incidence of depression for that age group. It should also be noted that the sample included subjects who were caregivers, stroke survivors, breast cancer survivors, and MI survivors, who are often older. The majority of the subjects were 49 and older. * 7. 101 subjects or 84.2% of the sample were 49 years of age or older. Add 35 + 37 + 29 = 101. Percent is found by 101 ÷ 120 = 0.8417 × 100 = 84.2%. Or the percentages in Table 1 can be added together: 29.2% + 30.8% + 24.2% = 84.2%. * 8. Yes, for this sample the ethnic categories are exhaustive because all 120 subjects fit into one of the ethnic categories of Caucasian, African American, or Asian. However, for another study, these three ethnic categories might not cover all of the different ethnic backgrounds of those subjects participating in the study. * 9. A median or middle score can be calculated for age and socioeconomic status, which means these variables are measured at the ordinal level. Since the variables are measured at the ordinal level, the categories can be rank ordered so that a middle score can be determined. The other variables are at the nominal level of measurement since they cannot be rank ordered. * 10. Nonparametric statistical analysis techniques are used to analyze nominal and ordinal level data. Nominal data can only be sorted into categories that are mutually exclusive and exhaustive and nonparametric analyses can be conducted on this level of data. Parametric analyses are usually conducted on variables that are measured at the interval or ratio level.
56 67
-------------------------------------------------
Top of Form
PageGo to the specified printed page number
Bottom of Form
Print Activate