(^110) Johns Hopkins Nursing Evidence-Based Practice: Model and Guidelines, Third Edition
central tendency (mean, median, and mode), are used to summarize these
differences.
Example: Nonexperimental Descriptive Comparative Design
Stickney, Ziniel, Brett, & Truog (2014) used a cross-sectional survey to compare
parental and healthcare providers’ attitudes and experiences of family participation
during unit rounds in a tertiary pediatric ICU. Parents and healthcare provider
questionnaires were developed in parallel to allow for direct comparison of their
perceptions of selected dimensions of rounds. Researchers administered the
questionnaires (surveys) to a convenience sample (nonrandom) of parents and
healthcare providers and described differences in responses between the two
groups.Descriptive correlational designs seek to describe relationships among variables.
Again, no attempt is made to understand causal relationships. The investigator
gathers information on at least two variables and conducts a statistical correla-
tion analysis between the two variables of interest to obtain a correlation
coefficient—a number ranging from –1 to 1. The correlation coefficient tells the
direction of the association between the two variables. If the correlation coef-
ficient is between 0 and 1, the correlation is positive: As one variable of interest
increases, so does the second variable. A negative correlation is depicted by corre-
lation coefficients between –1 and 0: As one variable increases, the other variable
decreases. The correlation coefficient also tells the reader the strength or magni-
tude of the correlation—that is, the closer this coefficient is to 1 (if positive) or –1
(if negative), the stronger the association between the two variables.Example: Nonexperimental Descriptive Correlational Design
Dabney and Kalisch (2015) examined patient reports of missed care and levels of
nurse staffing. The researchers used Pearson Correlation Coefficients to describe
the relationship between missed care (basic care, communication, timeliness) and
staffing levels (RN hours per patient day [HPPD], Nursing HPPD, RN skill mix).
Lack of timeliness was negatively associated with total Nursing HPPD (r = –0.09),
RNHPPD (r = –0.14) and RN skill mix (r = –0.13).