Characteristics Associated With School Nurse Childhood Obesity Prevention Practices

Susan B. Quelly


Pediatr Nurs. 2017;43(4):193-199. 

In This Article


This descriptive cross-sectional study was approved as exempt human research by the University of Central Florida's Institutional Review Board.


Recruitment of a purposive sample of Florida school nurses (N = 171) was facilitated by the Florida Association of School Nurses (FASN), the Florida School Health Association (FSHA), and 12 Florida county school districts. Participants self-identified as meeting inclusion criteria that required being an RN and presently working as a school nurse in Florida, or were employed as a Florida school nurse during the 2010–2011 school year. A power analysis determined that a sample size of 159 was needed to detect a medium effect size (0.25) in an ANOVA with three groups, assuming an alpha of 0.05 and a power of 0.80. The correlational and t test data analyses conducted in this study required a smaller sample size.


Study participants completed a self-administered, anonymous online or paper survey. Test-retest reliability was evaluated from 13 participants who completed the online survey twice. School nurses were offered a $5 gift card as an incentive for participating in the survey.

Between January and March 2012, 60 school nurse participants were recruited at two FASN conferences from a paper recruitment-consent message provided to attendees. This message explained inclusion and exclusion criteria, and that continuing to fill out the survey would constitute consent to participate in the study.

This same recruitment-consent message was emailed by the FASN, FSHA, and 12 Florida county school districts to their school nurses during this time frame. The remaining 111 participants accessed the online survey through Survey Monkey, a secure Internet survey. A request was included at the end of the online survey for participants to forward the recruitment-consent message to other Florida school nurses who might be interested in participating in the study.


A 70-item survey administered in another study included 17 demographic items used to collect data pertaining to personal, professional, and job-related characteristics of the school nurse participants. This survey also included 19 items that comprised the two scales measuring school nurse engagement in child-level and school-level COP practices.

Characteristics. Open-ended demographic questions on the survey were used to determine the age, number of years of nursing and school nursing experience, and years in current position. Questions with multiple choice or dichotomous response options were used to collect data about other personal, professional, and job-related characteristics of school nurses.

Personal characteristics included race/ethnicity, BMI, and age. Professional characteristics included level of nursing education, professional memberships, special certification, and hours of COP education. Job-related characteristics included school setting, job description, number of students for whom the school nurse provided care, socioeconomic level of the school(s) based on the percentage of students qualifying for free or reduced lunches as a proxy, and grade levels of the school.

COP practice scales. The frequency with which school nurses performed child-level and school-level COP practices was measured using two existing scales with 9 and 10 items, respectively (Kubik et al., 2007). COP practices were measured using 4-point Likert-type scales (1 = never; 2 = rarely; 3 = sometimes; 4 = often) similar to those described previously. Internal reliability for both scales was satisfactory based on two studies (Kubik et al., 2007; Quelly, 2015) reporting Cronbach's alpha ranging from 0.83 to 0.87 for child-level and school-level practices. Stability reliability was supported for both COP practices scales based on acceptable test-retest reliability coefficients for the child-level (r= 0.78) and school-level scales (r = 0.77) (Quelly, 2015).

Data Analysis

The Statistical Package for the Social Sciences Version 19.0 was used to analyze data. An a priori level of statistical significance (p < 0.05) was established, unless reported differently. Descriptive statistics that included means, standard deviations, and frequencies were examined and used to analyze sample demographic characteristics. Mean scores for the two COP practice scales were computed for each case and calculated using only answered items for 15 cases missing less than 10% of data. A mean scale score from a group with a comparable characteristic (highest nursing degree) was imputed for one case missing all data on the school-level COP practices scale.

The school nurse age, years of nursing experience, years of school nursing experience, years in current position, number of students provided care by a school nurse, the percentage of students qualifying for free or reduced lunch in their school(s), and both child-level and school-level COP practices were analyzed as continuous data. All other characteristic variables were analyzed as nominal data. The characteristics, number of years in current position, and number of students provided care, were not normally distributed, but the correlation analyses can tolerate moderate violations in normality assumptions due to the sample size.

The research question was addressed by conducting ANOVA and t tests to determine significant differences between nominal demographic variables and COP practices. The relationship between continuous demographic characteristic variables and COP practices was analyzed with correlation coefficients.