DNP 679 Module 1 Post Immersion 1 Written Assignment: Descriptive Statistics on Insomnia Severity Data Example

Seventy-Five Adults and a Skewed Scale: Describing the Class's Insomnia Severity Index Data

Student Name

Edson College of Nursing and Health Innovation, Arizona State University

DNP 679: Biostatistics: Principles of Statistical Inference

Instructor Name

Month Day, Year

Seventy-Five Adults and a Skewed Scale: Describing the Class's Insomnia Severity Index Data

Purpose and Data

This write-up describes the data our cohort collected before Immersion 1. Each of 15 students administered the Insomnia Severity Index (ISI) to five adult volunteers and recorded age, sex and whether the person worked rotating or night shifts, giving a pooled file of 75 records. The ISI asks seven questions about the past two weeks, each rated 0 to 4, and sums them to a total of 0 to 28 (Bastien et al., 2001). The purpose here is description only: to summarize who answered and how their scores are distributed, not to test a hypothesis. The figures that follow are the hypothetical teaching values used for this sample.

Data Checks Before Analysis

Before summarizing, the pooled file was screened. Each record was checked for missing items, since an ISI total is only valid when all seven items are answered; two records with one blank item were returned to the students who collected them and completed, so no data are missing. Every item was confirmed to fall between 0 and 4 and every total between 0 and 28, and ages were checked for impossible values. Sex and shift work were coded as numbers with labels, so Intellectus would treat them as categories rather than quantities.

Levels of Measurement

Each variable was classified before any statistic was chosen, because the level of measurement decides which summaries are meaningful (Polit & Beck, 2021). Sex and shift work status are nominal: their categories have no order, so they are summarized with counts and percentages. ISI severity category (none, subthreshold, moderate, severe) is ordinal, since the categories are ordered but the distances between them are not equal. Age is a ratio variable with a true zero. The ISI total is a summed rating scale; strictly it is ordinal, but by convention it is treated as interval for analysis, which permits a mean and standard deviation as long as the shape of the distribution is reported alongside them.

Results

Of the 75 respondents, 49 (65.3%) were female and 26 (34.7%) male, and 20 (26.7%) worked rotating or night shifts. Ages ranged from 19 to 69 years, with a mean of 41.7 (SD = 12.4) and a median of 42, close enough to suggest a fairly symmetric age distribution.

ISI totals ranged from 2 to 22, with a median of 9 and an interquartile range of 7 to 13. The mean was 10.2 (SD = 4.5) and the mode 9. In Intellectus, the histogram showed most scores clustered between 5 and 14 with a tail toward higher values, and the skewness statistic was 0.47, a mild right skew. The box plot showed no values beyond the whiskers. Because the mean sits above the median, pulled by the few higher scores, the median and interquartile range are the better primary summary, with the mean reported for comparison with published studies.

Half of the sample fell in the subthreshold range, and 13 respondents (17.3%) scored 15 or higher, in the moderate or severe clinical range. Morin et al. (2011) found that a cutoff of 10 best separated insomnia cases from good sleepers in a community sample; by that threshold, 37 of our respondents would screen positive, though the cutoff was validated for screening, not diagnosis.

Research Design and Its Limits

These data come from a cross-sectional design with a convenience sample. Everyone was measured once, so the data describe a single moment and cannot show change or sequence. Respondents were family members, friends and coworkers of nursing students, so they are likely healthier, more educated and more likely to work in health care than adults in general, and the results cannot be generalized to a population. The shift work proportion, for example, almost certainly reflects who nursing students know. Fifteen different people administered the questionnaire, which adds variation in how instructions were given. These limits do not make the data useless; they make them suitable for practicing description and for generating questions, such as whether shift workers score higher, that a better designed study could answer.

Summary

The pooled class sample was mostly female, middle-aged and one quarter shift workers. Insomnia severity was mildly right-skewed, centered on the subthreshold range, with about one in six respondents in the clinical range. The median of 9 and interquartile range of 7 to 13 describe the typical respondent best.

References

Bastien, C. H., Vallieres, A., & Morin, C. M. (2001). Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Medicine, 2(4), 297-307. https://doi.org/10.1016/S1389-9457(00)00065-4

Morin, C. M., Belleville, G., Belanger, L., & Ivers, H. (2011). The Insomnia Severity Index: Psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep, 34(5), 601-608. https://doi.org/10.1093/sleep/34.5.601

Polit, D. F., & Beck, C. T. (2021). Nursing research: Generating and assessing evidence for nursing practice (11th ed.). Wolters Kluwer.