Mutlidimensional Growth in a Camp Setting
In STAT 8420, I researched an article that set out to measure Youth Development Outcomes of children in a Camp Setting. I was tasked to analyze the provided statistcal data and report on the methods used to measure the data provided in the article. This was the hardest class for me, as I'm sure it was for many. However, prior to this class I would read the introductions and final analysis of empirical articles and now, I don't mind reading about the methods because I actually understand them better that I had before.
Initial Post:
Thurber et al. (2007) set out to measure Youth Development Outcomes of children that have “Camp Experience.” In particular they studied the positive changes in four domains: Positive Identity, Social Skills, Physical & Thinking Skills, and Positive Values & Spirituality by using ten constructs that make up these domains. The Ten constructs were: Self Esteem, Independence, Leadership, Friendship Skills, Social Comfort, Peer Relationships, Adventure and Exploration, Environmental Awareness, Values and Decisions, and Spirituality. The author’s believed that children that participated in a week or more of an organized day or resident camp “would result in growth in all four domains, as reported by parents, camp staff, and children themselves.” The study was done based on the theories of prevention science and positive youth development, arguing that summer camps provide space to reduce unhealthy behaviors and promote and nurture positive youth development. All selected camps for the study are accredited by the American Camp Association and serve children ages 8 to 14. Thurber et al. wanted to maintain variability by choosing camps from multiple locations across the US with different sponsorship. Families from each camp were selected to participate in pre and post surveys, as well as a six month follow up.
All quantitative and qualitative information and responses from the surveys were entered into SPSS. As predicted before the study, campers whose self- reporting score was lowest before camp had the highest gains in all categories. Statistically significant growth was found in all of the major constructs that made up the four domains. The highest growth was in Adventure and Exploration. The only negative effect was under peer relationships. Statistical significance was determined at p < or = to .05.
Utilizing an analysis of variance, the authors also compared the mean change scores between religious and non-religious camps. It showed that campers at religiously affiliated camps experience a higher growth in “Spirituality” than those that attended non-religious camps (F 3, 2081) = 2.99; p < or = .05.
This study demonstrates that “accredited summer camps of at least a week’s duration may all provide, to some degree and for most children, the essential ingredients for positive youth development.” The study demonstrated that very few children responded that their camp experience was negative and most gains in the four domains showed during the six month follow up survey or maintained the same level.
Response Post:
Thurber et. Al (2007) wanted to develop a research initiative that measured youth development outcomes of camp experience and show evidence of multidimensional growth. The main areas of they were interested in looking at were to see if campers reported increases in four domains: Positive Identity, Social Skills, Physical & Thinking Skills, and Positive Values and Spirituality. Ultimately they wanted to determine whether or not children developed at camp, and in what ways. They explored this through a descriptive, longitudinal study that was non experimental in design.
Methods
Researchers took a sample of camps accredited through the American Camp Association that served normally developing children between the ages of 8-14. Because of the self-reporting component of the study, they wanted to maintain the youngest age at 8. They also looked specifically at ACA accredited camps because of the intense standards in place that the camps must follow. While this doesn’t create a perfectly homogenous sample, it does help eliminate poor quality camps. After the camps were chosen, a matrix was created to show: camp type (resident or day), Sponsorship (agency sponsored, non-profit sponsored, for profit sponsored, and religiously affiliated camps), gender served (co-ed, all boys, all girls), and session length (1 – 4+ weeks). Researchers created a stratified sample by calling various types of camps with each part of the matrix (example: a resident, non-profit sponsored, all girls, two week camp), in each of the four quadrants in the United States. After phone calls and workshops, ultimately information was gathered from 80 camps.
Families were invited to participate in the study through three rounds of data collection: pre-camp survey, post-camp survey, and a six month follow up survey. Parents and children filled out their own questionnaires respectively. At the end, data was collected from 5,279 parents and children during the pre-camp survey (given four weeks prior to camp); 3,395 during the post-camp survey (given to the camper on the last day of camp, and to the parent two weeks after camp) ; and 2,293 in the six month follow up. Demographic characteristics of the respondents stayed relatively the same with one exception, minority family participation dropped. The mean age of campers stayed constant at 11.1 years old with a standard deviation of 1.9.
On top of the questionnaires sent to each of the families, camp counselors were asked to complete observation check lists at the beginning and end of camp. Data was received from all camps.
For this study, there was no tool in existence to measure the four domains and ten constructs simultaneously, so the researchers created one. The questionnaire is named the Camper Growth Index – Child Form (CGI – C). Another was created for parents – the Camper Growth Index – Parent Form (CGI – P). The surveys included 52 items/questions measured/answered by a 4 point Likert Scale. A factor analysis of the questionnaire showed 10 reliable constructs of 4-6 items each that fit into the four domains in which they wished to measure growth. The ten constructs, along with their alphas: Self Esteem ( ∝ = .75), Independence (∝ = .63), Leadership (∝ = .77), Friendship Skills (∝ = .69), Social Comfort (∝ = .66), Peer Relationships (∝ = .71), Adventure and Exploration (∝ = .66), Environmental Awareness (∝ = .76), Values and Decision Making (∝ = .76), and Spirituality (∝ = .81).
Validity of the data for the scale was established in three ways: 1. A factor analysis of the CGI – C and CGI – P. 2. Examination and revision by an expert panel. 3. Correlations with existing and established measures of constructs such as “self esteem.”
Scoring for the surveys included the scores from the CGI-C and CGI-P. Researchers averaged the scores of the items (from the Likert Scale) in the four domains and also averaged the scores of the items in the 10 constructs. This resulted in a mean score range from one to four.
In comparing data from Pre – Camp to Post – Camp, statistical significance is determined at p ≤ .05. The children report statistically significant growth in seven of the ten constructs. The largest growth was in the Adventure and Exploration construct. The only negative statistic was in peer relationships. Campers did not report growth in this area, in fact it declined. Parents post camp reports on their children showed similar growth, but in all ten constructs. The largest growth also being in Adventure and Exploration.
Comparing data from Pre – Camp to Six Month Follow Up, statistical significance is determined at p ≤ .05. Campers reported maintaining their gains in most categories except Making Friends, Adventure and Exploration, Values & Decisions, Environmental Awareness, and Spirituality. The mean of this data from the self reports went back to levels reported pre- camp. Parents again reported similarly, with Adventure and Exploration, Making Friends, and Environmental Awareness making statistically significant regressions.
“Exploratory correlation analyses and curve estimations of children’s self-reported precamp scores on the 10 constructs of the CGI-C suggested some small linear relationships – both positive and negative – with chronological age.” All campers that completed the pre and post camp surveys = precamp scores + (B x .1667). B is the unstandardized regression coefficient for age and .1667 is one – sixth of a year, the mean pre-post interval. “Paired-sample t-tests revealed that actual postcamp scores were significantly larger that estimated in most cases.” This suggests and shows evidence that children matured above and beyond what was expected during the pre-post survey interval. This may likely contribute the child’s development.
While their initial hypothesis was growth in all areas across various camps and locations, they also utilized the data to look at correlates of change, specifically: 1. Longer camp stays affected greater changes; 2. Intentionality emphasizing a specific aspect of development affects greater change in that area; 3. Children that had more room for growth would report the greatest gains; and 4. Children who enjoyed the camp the most, grew the most developmentally.
Bivariate correlational analyses in regard to days spent at camp showed very little correlation. In fact, the study showed that longer sessions strained peer relations and campers reported less interest in exploration and new activities (r= -.04, p ≤ .05; r = -.06, p ≤ .001).
Utilizing an analysis of variance, the researchers also compared the mean raw change scores between religious and non-religious camps. It showed that campers at religiously affiliated camps experience a higher growth in “Spirituality” than those that attended non-religious camps (F 3, 2081) = 2.99; p ≤.05). Post surveys confirmed that camper reported spiritual growth at a religiously affiliated camp was greater than those at an agency affiliated camp (p ≤ .01).
As predicted before the study, campers whose self- reporting score was lowest before camp had the highest gains in all categories. Best demonstrated in the comparison between the average pre-camp and post-camp scores in five categories. Those who scored between 0 and 2.99 on the pre-survey had a mean post-survey change of .29. “Between 3 and 3.249, mean pre-post change was .10; between 3.25 and 3.49, mean pre-post change was -.02; and between 3.75 and 4, pre-post change was -.06.”
In campers post-camp questionnaire, they were asked to rate their overall experience from 0 – 10, 10 being excellent. “The bivariate correlation with total CGI-C pre-post change scores was .17 (p ≤ .01). This suggested a modest relationship, if any, between growth at camp and children’s enjoyment of the experience.”
