SWIS and Ethnicity

Do Schools Using SWIS Take Advantage of the "School Ethnicity Report"?

Claudia G. Vincent

November, 2008

Evidence exists that students from minority backgrounds are referred to the office for violations of behavioral expectations at a disproportionately higher rate than their white peers (Raffaele Mendez & Knoff, 2003; Skiba, Michael, Nardo, & Peterson, 2002; Skiba et al., under review). Although disproportionate referral patterns have been documented, it is unclear to what extent schools actively address this issue. The School-wide Information System (see swis.org, May et al., 2006) is a web-based system that allows schools to collect an array of data on office discipline referrals (ODR).  Schools gain access to SWIS through an annual subscription and a SWIS facilitator providing training in how to use SWIS. Schools that have adopted SWIS commonly use information about the type of problem behavior, time and location of its occurrence, the student receiving the referral, and the resulting administrative decision to review their disciplinary practices. In addition, SWIS users have the option to examine the pattern of office discipline referrals by student ethnicity. SWIS allows schools to record (a) their overall school enrollment by ethnicity and (b) the ethnicity of individual students who receive an ODR. Based on this information, SWIS can then generate a "School Ethnicity Report" summarizing ODR data into the following 3 graphs (Todd, Horner, Sampson, & Amedo, 2008):

  • Percentage of all enrolled students by ethnicity and percentage of referrals by ethnicity
  • Percentage of all enrolled students by ethnicity and percentage of students with at least one referral by ethnicity
  • Percentage of students within each ethnic group who have at least one referral

At the school level and the individual student level, SWIS users can choose from 8 ethnicity categories: (1) American Indian/Alaskan Native, (2) Asian, (3) African American, (4) Hispanic/Latino, (5) Pacific Islander/Native Hawaiian, (6) White, (7) Unknown, and (8) Not Listed. Entering school enrollment ethnicity information into SWIS is optional. SWIS users can de-activate the "ethnicity" feature for their school to hide the "enrollment by ethnicity" fields. Entering individual student ethnicity is required, but schools can choose "unknown" or "not listed" from a drop-down menu to complete the required student ethnicity field. However, since both school level and individual student level ethnicity information is required to generate each of the 3 graphs in the SWIS ethnicity report, only schools that record functional information for both sets of data can derive practical value from this SWIS feature. Given the current policy mandates holding schools accountable for equitable educational environments where all students can succeed (NCLB, IDEA), our goal was to examine to what extent schools using SWIS take advantage of its ability to record and aggregate ethnicity data.

Data Description and Descriptive Analyses

We examined ODR data collected via SWIS during academic years (AY) 2005-2006, 2006-2007, and 2007-2008. Datasets for each year included all schools that used SWIS for the entire academic year, had an enrollment larger than 0, and recorded school days per academic year as larger than 0. All ODRs generated by these schools were included (minor and major referrals; for definitions of referral categories seehttp://www.swis.org/index.php?page=
resources;rid=10100
). Table 1 below provides an overview of the three datasets.

Grade level Number of schools by AY Percent of schools by AY Number of referrals by AY Percent of referrals by AY
  0506 0607 0708 0506 0607 0708 0506 0607 0708 0506 0607 0708
Pre 3 2 2 .2 .1 .1 478 1363 542 .0004 .1 .0002
K-6 1060 1441 1910 62.4 64.9 65.9 493,309 674,739 843,464 37.5 39.4 39.6
6-9 334 431 507 19.7 19.4 17.5 458,824 617,664 699,363 34.9 36.1 32.8
9-12 108 137 195 6.4 6.2 6.7 212,848 244,083 365,744 16.2 14.3 17.2
K8-12 194 208 284 11.4 9.4 9.8 150,208 173,145 222,558 11.4 10.1 10.4
Totals 1699 2219 2898 100 100 100 1,315,667 1,710,994 2,131,701 100 100 100

To examine the extent to which SWIS users (a) recorded school level ethnicity information, (b) recorded individual student ethnicity, and (c) accessed the SWIS ethnicity report, we used SPSS 16.0 to generate the following descriptive information: 

  1. the percentage of schools recording school enrollment by ethnicity 
  2. the percentage of all ODRs given to students from specific ethnic backgrounds 
  3. the percentage of schools accessing the SWIS ethnicity report  

The criterion for recording school enrollment by ethnicity was completion of the "enrollment by ethnicity" fields. The criterion for ODRs given to students from specific ethnic backgrounds was availability of any student level ethnicity information other than "unknown" or "not listed." And the criterion for schools accessing the SWIS ethnicity report was 3 or more recorded clicks on the ethnicity report feature per academic year by schools who recorded school level as well as student level ethnicity information and were therefore able to generate ethnicity reports containing functional data. We first examined patterns of recorded ethnicity information in SWIS across school levels and academic years. Because we were also interested in examining if US states differ in their use of SWIS ethnicity information, we arranged the information described above by state for each academic year.

Results

Figures 1 and 2 below show the percentages of SWIS users who recorded their overall 
school enrollment by ethnicity (Fig 1) and who recorded individual students' ethnicity (Fig 2) across grade levels and academic years. Overall, it appears that the majority of schools recorded neither school level nor individual student level ethnicity information via SWIS.  Only about 30% of all SWIS users recorded school enrollment by ethnicity, while only slightly more (approximately 40%) recorded individual student level ethnicity information. Across school levels, more middle schools (grades 6-9) provided school-level ethnicity information than other schools, while more high schools (grades 9-12) provided student-level ethnicity information than other schools. 
It is important to consider that we did not know the extent to which the same individual schools were represented in all of the three academic years. Overall, the number of schools increased from year to year (see Table 1). However, some schools might have continued their use of SWIS from year to year, while others might have discontinued their use, while still others might have newly adopted SWIS during any given year. Therefore, changes across academic years cannot be interpreted as a function of continued use of SWIS, but merely as a feature of overall SWIS use.


Figure 1


Figure 2

Figure 3 below illustrates the use of the various SWIS ethnicity features across all states for each of the 3 academic years. To facilitate interpretation of the graphs, all percentages are calibrated on the base counts of schools using SWIS indicated in parentheses after each state abbreviation on the x-axis. For example, in 2005-2006, 163 schools in Oregon used SWIS. Of those 163 schools, 30% recorded their overall school enrollment by ethnicity; of all ODRs given in those 163 schools, 44% were given to students whose ethnicity was recorded; and  of those 163 schools, 10% accessed the SWIS ethnicity report containing functional data 3 or more times during the academic year. 
It is again difficult to discern clearly interpretable patterns across states or across years. It is important to consider that high percentages on all three measures are easier to obtain when the school base count is low. Given that, it is encouraging to note that some states with a relatively high count of SWIS users (e.g. CO, IL, MD, NC) showed high percentages of schools recording enrollment and student ethnicity as well as accessing the SWIS ethnicity report. It might be reasonable to assume that SWIS ethnicity information usage patterns vary with state accountability policies, as well as local SWIS trainings; however without having accurate measures of these variables, these assumptions cannot be substantiated based on the current datasets.




Figure 3: Percentage of SWIS users recording school-level ethnicity, recording student level ethnicity, and accessing the SWIS ethnicity report.

Discussion and Further Directions

The descriptive analyses of the current data sets showed that, overall, the ethnicity features of SWIS seem to be heavily under-utilized. Schools using SWIS appear to put little emphasis on recording the data necessary to examine their referral patterns for equity across ethnic categories. This low usage of the SWIS ethnicity features is surprising given the somewhat extensive evidence that schools' disciplinary practices often result in disproportionate referral rates for students of minority backgrounds. Schools' reluctance to use the tools to address this issue is likely due to a multitude of reasons, among which might be (a) insufficient data entry time combined with large numbers of referrals, (b) insufficient emphasis on the ethnicity reporting features of SWIS during SWIS trainings, (c) schools' lack of motivation to examine their ODR patterns across ethnic categories, or (d) schools' lack of direction on how to address disproportionate referral patterns should they be identified.      
No information was available on the type of discipline approach schools represented in the SWIS dataset practiced. Although it is assumed that many schools that adopt SWIS also adopt school-wide positive behavior support, we cannot make this assumption based on the data we examined. However, regardless of their approach to discipline, encouraging SWIS users to make regular use of the SWIS ethnicity feature might provide schools with important information that could be used to create (more) equitable learning environments.
Our purely descriptive analyses clearly indicate the need for further more detailed analyses of ethnicity data and their use patterns in relation to a number of variables, including (a) state accountability policies, (b) SWIS trainings and facilitator activities, (c) schools' varying ODR rates, (d) the degree of schools' ethnic heterogeneity, or (e) grade levels. Closer study of these various factors might shed light on why evidence of disproportionate referral rates of students from minority backgrounds persists.

References

May, S., Ard, W., Todd, A. W., Horner, R. H., Glasgow, A., Sugai, G., et al. (2006). School-wide information system. Eugene: Educational and Community Supports, University of Oregon.

Raffaele Mendez, L. M. & Knoff, H. M. (2003).  Who gets suspended from school and why: A demographic analysis of schools and disciplinary infractions in a large school district. Education and Treatment of Children, 26, 30-51.

Skiba, R.J., Horner, R.H., Chung, C., Rausch, M.K., Seth, M., & Tobin, T. (2008). Race is not neutral: A national investigation of African American and Latino disproportionality in school discipline. Manuscript submitted for publication.

Skiba, R.J., Michael, R.S., Nardo, A.C., & Peterson, R. (2002).  The color of discipline: Sources of racial and gender disproportionality in school punishment.  Urban Review, 34,317-342.

Todd, A., Horner, R.H.,  Sampson, N. & Amedo, M. (2008). SWIS user's manual 2008. Version 4.2. Eugene, Oregon: Educational and Community Supports.

Citation for this Research Brief

Vincent, C. G. (2008, November). Evaluation brief: Do schools using SWIS take advantage of the "school ethnicity report"? OSEP Technical Assistance Center on Positive Behavioral Interventions and Supports. Web site: http://pbis.org/evaluation/evaluation_briefs/default.aspx