In: Statistics and Probability
Use this to answer questions: Teachers in 1 middle school learned about the positive effects of writing praise notes to students, which is 1 component of a positive behavior support. The authors intended for this procedure to promote a positive school environment and reinforce the appropriate use of social skills. Also, the authors instructed the teachers to use a direct instruction model to teach social skills lessons during 1st-period classes and praise students when they effectively demonstrated these skills. The authors analyzed the data to determine whether students receiving praise notes were less likely to receive an office discipline referral (ODR). The data revealed a significant negative correlation between the number of praise notes and number of ODRs that students received, indicating that as praise notes increased, the rate of ODRs decreased. The authors provide several hypotheses for this relation.
EFFECTIVE SCHOOLWIDE MANAGEMENT of disruptive behaviors is an ongoing national concern (Lewis & Sugai, 1999; Scott, 2001; Turnbull et al., 2002). School violence, discipline, and safety have been among the top concerns for U.S. educators (American Federation of Teachers, 1995-1996; Elam, Rose, & Gallup, 1998; U.S. Department of Education, 1995, 2005). When addressing students with problem behaviors, many schools continue to rely on punitive strategies (e.g., office or administrative disciplinary interventions, suspensions, expulsions) that do little to create a safe and positive educational environment (Lewis & Garrison-Harrell, 1999). These types of interventions tend to be reactionary rather than preventive and proactive. In addition, these types of responses do little to teach new behaviors or to increase the likelihood that positive replacement behaviors would be used in the future (Knoff, 2003). Punitive disciplinary measures can certainly be one approach to behavior management, but if punishment is the only approach used, student behaviors are unlikely to change over the long term. When administrators and other school adults intentionally seek opportunities to build and strengthen adult-youth relationships, they may actually be decreasing the likelihood that students will act out in the future (Young, Black, Marchant, Mitchem, & West, 2000).
Method
Participants and Setting
Participants were 70 teachers (48 women, 22 men) and 1,809 sixth- and seventh-grade students (927 boys [51%], 882 girls [49%]; 86% Caucasian, 11% Hispanic, 1% Native American, and 1% Pacific Islander, African American, or Asian) at secondary schools in the western part of the United States. Approximately 39% of these students qualified for free or reduced-price lunch.
This school was in the 3rd year of implementing a schoolwide PBS model. A school planning committee-comprising school administration, selected teachers, and representatives from a local university-discussed concerns and developed schoolwide goals. School faculty and staff members addressed these goals by providing social skills lessons, instructing students on expectations for their behavior, and agreeing to increase positive feedback to students.
Procedure
We instructed the teachers that during this study, which was conducted across 2 consecutive school years, they were to write praise notes to students whose behavior exemplified schoolwide PBS goals. At the beginning of the school year, as a part of a 2-day PBS training sequence, teachers were taught how to effectively praise students. Teachers were given blank praise notes with instructions on how to fill them out.
Measures
Praise notes were printed in triplicate on no-carbonrequired paper. Students were given the original copy. Teachers turned in a copy for drawings and prizes; we used this copy for data analysis. Last, the third copy was given to parents during parent-teacher conferences. Praise note data (e.g., name of student, name of teacher, date, behavior for which the student was praised) were entered into a database. Fewer than 1% of notes were incomplete and therefore eliminated from the analyses.
We tracked students' ODRs using a district-maintained database and collected teacher-written praise notes for the 2005-2006 and 2006-2007 school years. Praise note and ODR data were analyzed quantitatively using SPSS statistical analysis software (Version 15.0). The unit of analysis was number of praise notes written per day per 100 students. This measure allowed for all months to be compared equally despite differences in number of days per month. It was also consistent in the event of changes in student body size. The unit of analysis for ODRs was also number of ODRs written per 100 students per day. We used bivariate correlations to examine the relation between total praise notes written and number of ODRs for each month.
Over the course of this 2-year study, 14,527 praise notes were written, and 2,143 ODRs were recorded (see Figures 1 and 2). There was a significant negative correlation between the total number of praise notes written to the student body and the number of ODRs for the student body (r = -.551, p < .05), indicating that, as praise notes increased, ODR rates decreased. In addition, for the subgroup of students who received at least one ODR, there was a significant negative correlation between praise notes received and number of ODRs: As praise notes increased among students with at least one ODR, their rates of ODR decreased (r = -.553, p < .05).
The general aim of this study was to explore how teachers' use of praise notes to students demonstrating competency with social skills would influence ODRs. The results showed that praise notes and ODRs had a significant negative correlation: As praise notes increased, rates of ODR decreased. Hence, the data provide some evidence that increasing teacher praise notes may have been influencing the decrease in ODRs. However, more closely controlled research is needed.
As with any descriptive research, the results of this study should be considered as correlational-not causal-relations. There are several variables that could have contributed to a decrease in ODRs: Social skill instruction may have been a sufficient intervention to decrease ODRs. Also, ODRs may have decreased as administrators and teachers became more skilled in responding to behaviors that led to ODRs. It is also possible that in noticing and praising positive student behavior, teachers may have overlooked or become less focused on inappropriate behaviors. Although the cause of lower ODR rates cannot be determined by this descriptive study, it appears that teacher praise contingent upon the use of social skills had positive outcomes for students and for the overall school climate-reinforcing positive behaviors and decreasing rates of ODR.
Questions:
What statistical test was used in "Using Teacher-Written Praise Notes to Promote a Positive Environment in a Middle School?"
Did the authors use the correct statistical test? In other words, what was their rationale for using this test (i.e., were the variables discrete or continuous and was the test appropriate for this type of data?)
What was the research question? How did the statistical test address and answer the research question?
How did the authors interpret the results of this study?
Suppose we conduct the following statistical experiment. We select a random sample of size n from a normal population, having a standard deviation equal to ?. We find that the standard deviation in our sample is equal to s. Given these data, we can define a statistic, called chi-square, using the following equation:
?2 = [ ( n - 1 ) * s2 ] / ?2
The distribution of the chi-square statistic is called the chi-square distribution. The chi-square distribution is defined by the following probability density function:
Y = Y0 * ( ?2 ) ( v/2 - 1 ) * e-?2 / 2
where Y0 is a constant that depends on the number of degrees of freedom, ?2 is the chi-square statistic, v = n - 1 is the number of degrees of freedom, and e is a constant equal to the base of the natural logarithm system (approximately 2.71828). Y0 is defined, so that the area under the chi-square curve is equal to one.
In the figure below, the red curve shows the distribution of chi-square values computed from all possible samples of size 3, where degrees of freedom is n - 1 = 3 - 1 = 2. Similarly, the green curve shows the distribution for samples of size 5 (degrees of freedom equal to 4); and the blue curve, for samples of size 11