Special Education Personnel Absences in Inclusion-Oriented Schools: Implications for Building Effective Service Delivery Models

Abstract

This study explored relationships between special education personnel absences and demographic, policy, and practice variables to identify potential actions that could increase access to qualified personnel and continuity of instruction. Findings from 51 inclusion-oriented schools indicated that special educators and special education paraprofessionals were absent 12 days per year on average. Special educator absences were correlated with variables amenable to action by school leaders including special educator school density (i.e., ratio of special educators in full-time equivalents to total school enrollment) and special services concentration (i.e., ratio of special educators to special education paraprofessionals in full-time equivalents). Special education paraprofessional absences were not correlated with these variables. Implications for practice and suggestions for future research are discussed.

Keywords

systems change legal/policy issues paraprofessionals special education access to education inclusion

Success in any endeavor starts with simply being present. In education, this idea is manifested by the significance placed on student attendance. Student attendance is key for school engagement (Christenson, Reschly, & Wylie, 2012) and a predictor of academic progress and achievement (Gottfried, 2010). However, the attendance of teachers has received less attention. Consistent access to instruction from highly qualified personnel and constructive student–teacher relationships are critical for student outcomes (Hattie, 2009) and are dependent on both students and teachers being present in school. Research on the teacher side of this equation has largely focused on shortages, turnover, and attrition (Billingsley, 2004; Simon & Johnson, 2015), but few studies have examined the importance of absences and none specifically for special education personnel. The purpose of this study is to address this gap in the literature by providing an initial exploration of special education teacher and paraprofessionals’ absences.

Effects and Scope of Teacher Absences

The importance of access to high-quality teachers for student achievement is emphasized in both general and special education research (Billingsley, 2011) and is a primary impetus for studies on teacher absences. Although the few studies on this topic have not specifically examined absences by special education personnel, they do support the hypothesis that teacher absences and student progress are connected. Significant relationships have been found across a range of grades and measures of progress in small rural samples (Woods & Montagno, 1997), large urban school districts (Herrmann & Rockoff, 2012; Miller, Murnane, & Willett, 2008), and statewide samples (Clotfelter, Ladd, & Vigdor, 2009; Ehrenberg, Ehrenberg, Rees, & Ehrenberg, 1991; Roby, 2013). Larger negative effects on achievement were found when examining absences that occurred prior to exams (Herrmann & Rockoff, 2012) and when absences were unexpected (Miller et al., 2008). The only experimental trial on this topic found that rural elementary schools in India that instituted a teacher incentive program had fewer teacher absences (42%–21%), 30% more instructional time, and student test scores that were 0.17 standard deviations higher than comparison schools (Duflo, Hanna, & Ryan, 2012).

Several studies have also examined the scope of teacher absences. Findings indicate teachers missed an average of 10–11 days per year, or approximately 6% of instructional days (Clotfelter et al., 2009; Ehrenberg et al., 1991; Herrmann & Rockoff, 2012; Joseph, Waymack, & Zielaski, 2014; Miller, 2008, 2012), compared to an average of 3% of days missed by U.S. workers with other full-time jobs (Miller, 2012). The prevalence and potential negative impact on quality instruction led the U.S. Department of Education’s (U.S. DOE) Office for Civil Rights to ask schools to report teacher absences (e.g., sick days and personal leave) in their biennial Civil Rights Data Collection survey. Schools reported that 27% of teachers were absent more than 10 days during the 2013–2014 school year; within schools, the percent of teachers missing more than 10 instructional days each year ranged from 0%–100% (U.S. DOE, 2016). Nationally, teachers characterized as chronically absent, those absent 18 or more days per year, were responsible for 34% of teacher absences (Joseph et al., 2014). These absences reflect a significant cost for schools, as hiring substitute teachers and related administrative costs were estimated at least $4 billion annually (Miller, 2012).

Predictors of Teacher Absences

A number of factors may contribute to teacher absenteeism, and we begin with the simplest explanation; teachers do get sick and appropriately take time off. In fact, a public health analysis showed that teachers’ work surfaces had the highest concentrations of bacteria compared to other professions (Hitti, 2006). In addition, teachers are in continuous close proximity to their students, who are more susceptible than adults to common acute respiratory illnesses (e.g., cold, flu; Monto, Malosh, Petrie, Thompson, & Ohmit, 2014). Higher levels of bacterial and viral contamination of work surfaces in schools have been shown to increase student absences (Bright, Boone, & Gerba, 2010).

Only a handful of studies have empirically examined predictors of teacher absences, and the findings do not provide clear or consistent guidance for how to reduce absenteeism. For example, schools with higher student enrollments have been associated with more teacher absences (Ehrenberg et al., 1991), but this relationship has not been consistent across samples (Bradley, Green, & Leeves, 2007). Other studies have indicated absences themselves are correlated, as teachers were more likely to be absent in schools in which other teachers and students had higher levels of absenteeism (Bradley et al., 2007; Ehrenberg et al., 1991; Herrmann & Rockoff, 2012; Miller et al., 2008). Some districts’ leave policies have been associated with more teacher absences (Miller, 2012), including the number of leave days available and sick leave banks (where teachers can accumulate unused sick leave, Ehrenberg et al., 1991). Conversely, certain district policies may reduce teacher absences, such as: (a) buyout of unused sick leave at retirement (Ehrenberg et al., 1991; Joseph et al., 2014), (b) teachers paying for sick days beyond those allotted (Clotfelter et al., 2009), and (c) providing additional compensation or leave for better attendance (Duflo et al., 2012; Joseph et al., 2014). Importantly, none of these studies examined predictors for special education personnel absences.

Special Education Service Delivery as Starting Point

If a general goal is to increase special educator attendance to improve access for students with disabilities to consistent instruction from highly qualified personnel, then a starting point can be examining research identifying malleable factors for potential predictors of special education personnel absences. First, research on working conditions in schools has shown that the organizational structures, supports, and resources available to teachers can promote quality instruction (Bettini, Crockett, Brownell, & Merrill, 2016). However, when problematic working conditions exist (e.g., role ambiguity, role conflict), it can lead to burnout (Brunsting, Sreckovic, & Lane, 2014; Garwood, Werts, Varghese, & Gosey, 2017). Teacher absences have been considered a potential indicator of teacher stress and burnout (Maslach & Leiter, 2005). In a survey of school administrators, nearly half of long-term absences by teachers were attributed to stress or related conditions (Bowers, 2004). While teacher turnover and burnout may be higher in high-poverty schools serving more students of color (Simon & Johnson, 2015) or rural students (Garwood et al., 2017), this may be a result of problematic working conditions rather than student demographics.

A second line of research has explored strategies for improving special education service delivery in schools supporting inclusion for students with disabilities (Giangreco, Broer, & Suter, 2011; Giangreco, Doyle, & Suter, 2012; Giangreco, Suter, & Hurley, 2013; Suter & Giangreco, 2009). Such inclusion-oriented schools seek to provide appropriate educational opportunities and levels of support to students with the full range of disability characteristics in chronologically age-appropriate, general education classes those students would attend if they were not labeled disabled (Giangreco & Suter, 2015). In these studies, participating schools exceeded the national average for the percentage of students with disabilities placed in regular classrooms 80% of the time or more (64%; U.S. DOE, 2018a), and with no more than 1% of students with disabilities sent out-of-district. While based on the Individuals with Disabilities Education Act (IDEA) and its least restrictive environment provisions, inclusion-oriented refers not solely to where a student receives instruction, but also how. Inclusion-oriented efforts are designed to ensure access to (a) the general education curriculum taught by highly qualified content-area teachers with the support of special educators, (b) peers without disabilities, and (c) other valuable opportunities often associated with general education classrooms (e.g., higher expectations, incidental learning, reduced stigma associated with segregated settings, opportunities for friendship development, and self-determination). The term inclusion-oriented is used rather than inclusive because few schools embody all elements under all circumstances that might be characterized as inclusive. While inclusion-oriented schools may engage some practices that could be questioned as to their alignment with inclusivity, they typically are aware of these areas in need of improvement and are striving to address them.

Research has identified two malleable indicators that hold the potential to enhance inclusive special education service delivery: (a) special educator school density (i.e., number of special educator full-time equivalents [FTE] per total enrollment) and (b) special services concentration (i.e., ratio of special educator FTE to special education paraprofessional FTE; Giangreco et al., 2013). While other research has examined caseload size (e.g., ratio of special educators to students receiving special education) and types of students served (Russ, Chiang, Rylance, & Bongers, 2001), the variations across special educators and caseloads are so variable and complex (e.g., percentage of time assigned to students, range of grades supported, number of paraprofessionals supervised) such that accurate comparisons are challenging. Resource reallocation (e.g., trading in three paraprofessional positions to hire one special educator) is an example of a cost-neutral action schools can take to improve both special educator school density and special services concentration (Giangreco et al., 2011). In previous research, these variables were correlated with special educators’ self-efficacy ratings based on work responsibilities (Giangreco et al., 2013).

Purpose and Research Questions

The current study represents an initial pilot investigation of special education teacher and special education paraprofessional absences in a convenience sample of inclusion-oriented schools. This study begins to fill a gap in the literature by providing data where none currently exists. It is important because special education personnel absences may interfere with students’ continuity of instruction and access to qualified personnel who know them. Better understanding the relationships between absences and variables over which schools have control (e.g., personnel deployment) may contribute to more effective inclusion-oriented service delivery. Our study had two research questions:

Research Question 1: What are the absence rates of, and potential differences between, special education teachers and special education paraprofessionals?

Research Question 2: How are special education personnel absence rates correlated with school demographic and special education personnel variables?

Method

Design

The current investigation employed a cross-sectional quantitative design to examine special education personnel absences and their relationship to school variables. Data were collected for three consecutive school years (2010–2011, 2011–2012, and 2012–2013) from a convenience sample of schools in Vermont, a predominantly rural northeastern state. These schools were chosen because they had previously participated in data-collection efforts with the study authors and were considered inclusion-oriented as described in the introduction.

Procedures

An Institutional Review Board–approved lay summary (e.g., study purpose, research questions, and description of the data sought), including a link to a secure online questionnaire where data could be submitted, was emailed to school administrators. The lay summary emphasized that we were not requesting any personally identifiable information about employee absences but rather only school-level data and aggregated absence numbers. Administrators also were supplied with a separate data-collection worksheet that paralleled the online questionnaire, so they could collect the data prior to logging on for data entry.

Participating Schools

Data were requested from 90 public schools in Vermont (no schools from our previous data collection were excluded from recruitment). Initially, 68 schools agreed to participate, and ultimately, 51 submitted data (response rate 56.66%). School administrators who did not participate cited lack of time and difficulty gathering the requested data as primary reasons. The 51 schools were in 15 different supervisory unions or districts representing all grade levels: elementary (n = 24), elementary-middle (n = 13), middle (n = 7), middle-high (n = 3), and high school (n = 4). Schools were situated in every region of the state in areas designated as rural (n = 29), town (n = 13), and suburb (n = 9); and no schools were in cities (U.S. DOE, 2018b).

On average, schools enrolled 331 students (SD = 246), with total enrollments ranging from 22–1,134. Students were identified primarily as White and non-Hispanic, with schools ranging from 73%–100% (M = 92.46%, SD = 5.03) of students in both groups. The percentage of students eligible for free or reduced-price lunch had a wider range across schools: 10%–78% (M = 35.92%, SD = 15.35), although most schools (82%) were designated as low (25% or less free or reduced-price lunch, n = 13) or mid-low (25%–50% free or reduced-price lunch, n = 29) poverty schools. Only one school was considered high-poverty with more than 75% free or reduced-price lunch (U.S. DOE, 2018b).

Measures

A questionnaire was created for this study with items informed primarily by two previous studies (Giangreco et al., 2013; Suter & Giangreco, 2009) and an earlier review of national data (Giangreco, Hurley, & Suter, 2009). For consistency, respondents were provided with definitions of terms in the questionnaire (see below). After reading the definitions, participants responded to each item for each of three consecutive school years in an effort to present a more accurate representation over time. If items were missing, or changed greatly from year to year, study authors followed up with respondents to confirm accuracy. Some school demographic data were accessed from the Common Core of Data provided by the National Center for Education Statistics (NCES) matched to the three study school years (U.S. DOE, 2018b).

Total school enrollment

School administrators reported the total number of students with and without disabilities served in each school.

Race and ethnicity

We obtained student race and ethnicity data from NCES for our participating schools matched to the three study school years (U.S. DOE, 2018b).

Free and reduced-price lunch

To create an indicator of poverty, we used NCES data matched to the three study school years (U.S. DOE, 2018b). The percentage of students with free and reduced-price lunch was calculated by dividing students with free and reduced-price lunch by total enrollment.

Special education personnel FTE

Schools were asked to count special educators and special education paraprofessionals in FTE for each full year. For example, if the school year began with three FTE special educators and a fourth was hired full-time in the middle of the year, the cumulative for the year would be 3.5 FTE.

Special educator FTE

These include all licensed special education teachers (in FTE) functioning in the role of a special educator. Therefore, it does not include individuals who are licensed special educators but functioning in other roles (e.g., paraprofessional, classroom teacher, and administrator). Schools were asked not to include speech/language pathologists or other related services providers (e.g., physical therapist, occupational therapist, and school psychologist).

Special education paraprofessional FTE

These include all special education paraprofessionals (in FTE), regardless of title, including those hired by the school district or supervisory union and through a contracted arrangement (e.g., interventionists/assistants from other agencies). Paraprofessionals are those who must be supervised by a licensed professional in their assigned role, even though some might have a license. General education paraprofessionals were not included. In cases where an individual has a split assignment between both special and general education, only the portion of the FTE associated with special education was counted. Schools were also asked to report on the subsets of special education paraprofessional FTE assigned to students one-to-one and special education paraprofessional FTE hired as contracted employees through external agencies.

Special educator school density

This ratio was calculated by dividing total school enrollment by special educator FTE for each school.

Special services concentration

This ratio was calculated by dividing special education paraprofessional FTE by special educator FTE for each school.

Sick and personal days available

Schools were asked to report on the number of sick and personal days available for paid leave. These numbers were reported separately for special educators and special education paraprofessionals.

Days absent

Absences were defined as days missed due to illness, family illness or bereavement, personal days, jury duty, or any other reason that a special educator or special education paraprofessional is temporarily absent from school for nonwork reasons. Consistent with the federal definition (U.S. DOE, 2016), teacher absenteeism does not include absences for work-related purposes (e.g., professional development) or long-term leave (e.g., parental leave, military service, and extended medical leave). Long-term absences may be covered in the personnel FTE data (e.g., if the person is replaced). Total days absent were reported separately for special educators and special education paraprofessionals and used to calculate: (a) Average special educator absences (number of days special educators were absent divided by special educator FTE) and (b) average special education paraprofessional absences (number of days special education paraprofessionals were absent divided by special education paraprofessional FTE).

Percent of occurrences the school employed substitutes

Schools were asked to estimate the percent of occurrences substitutes were hired separately for special educator and special education paraprofessional absences. A substitute was defined as an individual who does not work at the school daily and is hired on a short-term temporary basis to cover the daily duties of the absent employee; it does not include school personnel who work in the school and are temporarily reassigned to partially or fully cover the duties of the absent employee.

Descriptive Data

Table 1 summarizes descriptive data for study variables. On average, the schools employed 4.56 (SD = 3.89) special education teachers in FTE and 12.77 (SD = 10.64) special education paraprofessionals FTE. The schools directly employed most special education paraprofessionals (95%), while approximately 5% (M = 0.7, SD = 1.14) were contracted to work in the schools through outside organizations (e.g., regional mental health agencies). About 42% (M = 5.38, SD = 5.17) of special education paraprofessionals in FTE were assigned in one-to-one arrangements to support individual students.

Table 1.

School Demographics and Descriptive Data for Study Variables (N = 51).

Variable	M	SD	Range	Skew
Total school enrollment	331.57	245.95	22–1,134	1.29
Percent students identified as:
White	92.46	5.03	72.9–100.0	−1.68
Two or more races	2.51	3.58	0.0–21.5	3.83
Hispanic	1.70	1.37	0.0–6.0	1.26
Asian	1.46	1.30	0.0–4.6	0.76
Black or African American	1.34	1.06	0.0–5.0	1.03
Percent students free and reduced-price lunch	35.92	15.35	10.2–78.1	0.59
Special educator FTE	4.56	3.89	0.8–19.2	1.87
Paraprofessional FTE	12.77	10.64	1.3–56.4	1.81
Special educator school density	80.00	25.16	36.3–159.1	0.84
Special services concentration	3.03	1.36	0.6–8.0	1.24
Special educator sick and personal days^a	20.01	3.43	15.0–28.0	1.06
Paraprofessional sick and personal days^a	14.82	3.14	9.0–19.0	−0.45
Average special educator absences	12.21	5.02	3.5–26.8	0.61
Average paraprofessional absences	11.56	4.54	2.8–21.2	−0.15
Percent occurrences substitute employed:
Paraprofessional absences	81.02	26.20	0.0–100.0	−1.68
Special educator absences	56.29	36.49	0.0–100.0	−0.33

Note. FTE = full-time equivalents.

Special educator and paraprofessional sick and personal days refer to number of sick and personal days available.

Special educator school density (i.e., the ratio of special educator FTE to total school enrollment) averaged 1:80.00 (SD = 25.16) and ranged substantially from approximately 1:36 (one special education teacher FTE for every 36 students attending the school) to approximately 1:159 (one special education teacher FTE for every 159 students attending the school). The special services concentration (i.e., the ratio of special education teacher FTE to special education paraprofessional FTE) averaged 1:3.03 (SD = 1.36), meaning schools employed three times as many special education paraprofessionals as special educators in FTE. Special services concentration ranged from 1:0.61 (one special education teacher FTE for every 0.61 special education paraprofessional FTE) to 1:8.00 (a school that employed eight times as many special education paraprofessionals as special education teachers in FTE). Notably, all but one school (n = 50; 98%) employed more special education paraprofessionals than special education teachers in FTE (i.e., special services concentration greater than 1:1).

Most schools offered a higher combined number of paid medical/personal days (over 5 additional days) to their teachers and special educators (M = 20.01 days, SD = 3.43) than they did to their paraprofessionals (M = 14.82, SD = 3.14). In only one school system was the number of paid medical/personal days the same for teachers/special educators and paraprofessionals. Allowable accumulation of sick/medical days varied widely. Special educators held salaried positions while paraprofessionals receive hourly pay. Approximately 80% of teachers in this state belonged to a labor union (ranked 24th in the United States; Winkler, Scull, & Zeehandelaar, 2012). During the three years reflected in the data, only during 2010 did the state report a shortage of special educators (U.S. DOE, 2017).

Data Analyses

Descriptive quantitative statistics, Pearson correlations, and standard multiple regression analyses were used for data analyses. Data collected for each of the three school years were averaged to create single variables. For example, the three total enrollment numbers provided by a school were averaged to create a single total enrollment variable for that school. Dependent variables included (a) average special educator absences (number of days special educators were absent divided by special educator FTE) and (b) average special education paraprofessional absences (number of days special education paraprofessionals were absent divided by special education paraprofessional FTE). Independent variables were tested in a series of three standard multiple regression models (a) total enrollment, percentage of students with free and reduced-price lunch, and available sick/personal days as indicators identified from previous literature on working conditions and teacher absences, (b) special education school density, special services concentration, and average special education paraprofessional absences (for special educator models) and average special educator absences (for paraprofessional models), and (c) a final model with significant variables identified from the first two models.

All data were evaluated using SPSS 25 for completeness and accuracy, then checked for violations of assumptions for multiple regression prior to conducting statistical analyses. No missing data or outliers were detected, residuals were normally distributed, and no evidence was found for nonlinearity, heteroscedasticity, or multicollinearity. The software program G*Power (Faul, Erdfelder, Buchner, & Lang, 2009) was used for power analysis indicating our sample size of 51 (with six predictors) provided 86% power to detect large effects (f² = 0.35, Cohen, 1988). This aligned with recommendations to have sample sizes of at least 50 for multiple regression (VanVoorhis & Morgan, 2007). Semi-partial correlations (sr) were included as indicators of predictor effect size. They represent the correlation of the dependent variable with the portion of each independent variable unique from the other predictors (Tabachnick & Fidell, 2013).

Results

Special Education Personnel Absences

On average, special educators were absent slightly more often (12.21 days per year, SD = 5.02) than special education paraprofessionals (11.56 days per year, SD = 4.54). While these numbers are relatively equivalent, they do not account for the reality that there are substantially more special education paraprofessionals in these schools than special educators. At the school level, special educators cumulatively missed an average of 57.18 school days per year (SD = 55.74) while special education paraprofessionals cumulatively missed an average of 158.51 days per year (SD = 146.40), totaling nearly 216 special education personnel days lost annually.

Based on varying local teacher and paraprofessional contracts (where teachers typically have more paid medical/personal days), on average paraprofessionals were absent 80.09% of available paid medical and personal days (SD = 34.12%) while on average special educators were absent 60.67% of available paid medical and personal days (SD = 21.66%). Schools reported hiring substitutes more often when special education paraprofessionals were absent (M = 81.02% of the time, SD = 26.20%) than when special educators were absent (56.29% of the time, SD = 36.49%). Hiring substitutes varied widely across schools ranging from 0%–100% of the time for both special educator and special education paraprofessional absences.

Relationship Between Special Education Personnel Absences and Independent Variables

Special educator absences

As seen in Table 2, average special educator absences were found to have significant and positive correlations with each independent variable except percentage of students with free and reduced-price lunch (r = −.15) and number of sick and personal days available for paraprofessionals (r = .05). A medium correlation was found with total enrollment (r = .31), and large correlations were found for (a) paraprofessional absences (r = .47), (b) number of sick and personal days available for special educators (r = .54), (c) special educator school density (r = .51), and (d) special services concentration (r = .54).

Table 2.

Pearson Correlations for Study Variables (N = 51).

Variable	1	2	3	4	5	6	7	8
1. Average special educator absences	—
2. Average paraprofessional absences	.47**	—
3. Total school enrollment	.31*	.35*	—
4. Percent students FRPL	−.15	−.07	−.26	—
5. Special educator sick and personal days^a	.54**	.17	.18	−.43**	—
6. Paraprofessional sick and personal days^a	.05	.17	.16	−.11	.14	—
7. Special educator school density	.51**	.25	−.03	−.30*	.46**	−.03	—
8. Special services concentration	.54**	.18	−.04	−.04	.33*	−.05	.59**	—

Note. FRPL = free and reduced-price lunch.

Special educator and paraprofessional sick and personal days refer to number of sick and personal days available.

p < .05. **p < .01.

Because special educator school density and special services concentrations are ratios, further explanation is warranted. In schools with a greater density of special educators (more special educator FTE per total enrollment), special educator absences tended to be lower. As density of special educators in a school became thinner, the average special educator absences increased. For example, in a school with a special educator school density ratio of 1:72 (one special educator FTE for every 72 students in the school), on average, the special educators were absent 10.4 days per year; whereas, in a school with a ratio of 1:110, on average the special educators were absent 20.3 days per year. Similarly, in schools where the special services concentration included ratios closer to the same number of special educators and special education paraprofessionals, special educator absences tended to be lower. As the special services concentration shifted toward increasingly more special education paraprofessionals than special educators, special educator absences increased. For example, in a school with a special services concentration ratio of 1:2 (twice as many paraprofessionals as special educators), on average, the special educators were absent 3.5 days per year; whereas, in a school with a ratio of 1:4.8 (nearly five times as many paraprofessionals as special educators), on average, the special educators were absent 19.5 days per year.

Three standard multiple regression models were run with average special educator absences as the dependent variable (see Table 3). The first model included three independent variables: Total enrollment, percent of students with free and reduced-price lunch, and available sick/personal days for special educators. The model accounted for 36% of the variance, R² = .36, F (3, 47) = 8.75, p < .001. Available sick and personal days was the only significant predictor and had a relatively large effect size (sr = .51).

Table 3.

Multiple Regression Results for Average Special Educator Absences (N = 51).

Variable	Model 1			Model 2			Model 3
Variable	B	SE	sr	B	SE	sr	B	SE	sr
School enrollment	0.01	.002	.25
Percent students with FRPL	0.05	.043	.14
Available sick and personal days	0.82**	.190	.51				0.53**	.153	.34
Absences by paraprofessionals				0.39**	.122	.34	0.38**	.111	.34
Special education school density				0.04	.027	.17
Special services concentration				1.27*	.487	.28	1.31**	.386	.33
R ²	.36			.46			.55
Adjusted R²	.32			.43			.52
F	8.75**			13.59**			19.02**

Note. sr = semi-partial correlation; FRPL = free and reduced-price lunch.

p < .05. **p < .01.

The second model included average paraprofessional absences, special educator school density, and special services concentration. The model was significant, F (3, 47) = 13.59, p < .001, predicting 46% of the variability in average special educator absences. Average absences by paraprofessionals (sr = .34) and special services concentration (sr = .28) each significantly contributed to the model with medium effect sizes, while special educator school density did not.

In the third model, we included independent variables that significantly predicted average special educator absences from the previous two models. The model was significant, F (3, 47) = 19.02, p < .001, predicting 55% of average special educator absences. Each of the three independent variables were significant with medium effect sizes: (a) available sick/personal days for special educators (sr = .34), (b) average absences by paraprofessionals (sr = .34), and (c) special services concentration (sr = .33).

Paraprofessional absences

Pearson correlations (see Table 2) revealed only two significant correlations for average paraprofessional absences: Average special educator absences (r = .47) and total enrollment (r = .35). Percent of students with free and reduced-price lunch, available sick days for paraprofessionals, special educator school density, and special services concentration did not show significant correlations with average paraprofessional absences. Three standard multiple regression models were run with average paraprofessional absences as the dependent variable (see Table 4), which revealed similar patterns of relationships as the Pearson correlations. Results showed smaller R² values than the models predicting average special educator absences. Only the third model accounted for more than 25% of the variability in average paraprofessional absences with the predictors total enrollment (sr = .21) and average special educator absences (sr = .38), F (2, 48) = 8.71, p = .001.

Table 4.

Multiple Regression Models for Average Paraprofessional Absences (N = 51).

Variable	Model 1			Model 2			Model 3
Variable	B	SE	sr	B	SE	sr	B	SE	sr
School enrollment	0.01*	.003	.32				0.004	.002	.21
Percent students with FRPL	0.01	.042	.03
Available sick and personal days	0.17	.200	.12
Absences by special educators				0.46**	.143	.41	0.36**	.118	.38
Special education school density				0.01	.030	.06
Special services concentration				−0.45	.560	−.10
R ²	.13			.23			.27
Adjusted R²	.08			.18			.24
F	2.43			4.76**			8.71**

Note. sr = semi-partial correlation; FRPL = free and reduced-price lunch.

p < .05. **p < .01.

Discussion

Access to quality teachers and effective instruction is critical for students (Hattie, 2009), particularly students with disabilities (Billingsley, 2011). Similarly, research on inclusion-oriented schools has shown how schools can change their special education personnel utilization to provide better access to highly qualified teachers and support inclusive educational opportunities (Giangreco & Suter, 2015). These efforts are likely at risk when personnel are unexpectedly absent, and a growing body of research has shown connections between teacher absenteeism and student progress (Clotfelter et al., 2009; Ehrenberg et al., 1991; Herrmann & Rockoff, 2012; Miller et al., 2008). This study represents the first examination of special educator and special education paraprofessional absences in the literature. Results from this pilot study revealed several important findings related to special education personnel absence rates and predictors representing school demographic and special education personnel variables.

Special Education Personnel Absence Rates

While some level of illness-related absence is expected, given the nature of schools, it is notable that the special educators and paraprofessionals in this sample were absent, on average, approximately 12 days per year (approximately 2.5 work weeks). These reported special education personnel absences add up. On average, approximately 216 special education personnel days were lost annually. While these absences appear to be only 1–2 days more per year than the reported estimates for teachers (10–11 days absent per year; Joseph et al., 2014; Miller, 2012), the difference is actually larger. Previous research included personal and family leave as part of teacher absences, whereas this study excluded such long-term leave, where a replacement would be hired. If extended leaves (e.g., family and military) were included in this study, the days absent would exceed our reported number, but by how much is unknown. The impact of absences also adds up when considering the costs of hiring substitutes 81% of the time for special education paraprofessional absences and 56% of the time when special educators are absent. Exploring why schools were more inclined to hire substitutes for special education paraprofessionals raises an important question for future research.

Predictors of Special Education Personnel Absences

Findings suggest special educator absences were linked to several, but not all, of the school-level predictors included in this study. One of the strongest predictors was available sick/personal days for special educators; with each additional day available, special educators were predicted to be absent half a day (holding other predictors constant, see Table 3, Model 3). This finding is consistent with previous research that indicated a link between teacher absences and district leave policies (Ehrenberg et al., 1991; Miller, 2012). Another significant predicator was special education paraprofessional absences, with special educators expected to miss nearly two days of school for every five days that special education paraprofessionals are absent (holding other predictors constant). Earlier studies also found higher teacher absenteeism in schools where other teachers and students were absent more often (Bradley et al., 2007; Ehrenberg et al., 1991; Herrmann & Rockoff, 2012; Miller et al., 2008).

Special services concentration was a significant predictor of special educator absences, but the relationship with special educator school density was less clear. For special services concentration, the final regression model (see Table 3) indicated that for every additional paraprofessional (FTE) a school adds (relative to special educator FTE and holding the other predictors constant), special educators are expected to be absent an additional 1.3 days per year. Special educator school density had a strong correlation with average special educator absences (see Table 2), but almost no relationship in the regression analyses when special services concentration and average special education paraprofessional absences were included. Rather than suggesting special educator school density is unimportant, it may indicate its relationship to average special educator absences is mediated by these other variables. Alternately, this finding could represent a measurement artifact as special educator school density and special services concentration each include special educator FTE in their calculation. Therefore, it is important to consider each of these malleable variables independently and in combination.

Challenging working conditions, stress, and burnout (Brunsting et al., 2014; Maslach & Leiter, 2005) are plausible explanations for higher rates of special educator absences in schools. The demographic variable included as an indirect indicator of school working conditions was the percent of students with free and reduced-price lunch, but it was not significantly related to special education personnel absences. A school’s percent of students with free and reduced-price lunch was significantly and negatively correlated with available sick and personal days for special educators (see Table 2). However, this study included mostly low or mid-low poverty schools, so examining these relationships with samples including high-poverty schools must be left to future studies. In addition, it will be important for future research to explore relationships between special services concentration, special educator school density, and special educator working conditions. For example, the average special services concentration in this study was three times as many special education paraprofessionals as special educators and ranged as high as eight times as many. With such paraprofessional-heavy special services concentration, special educators could spend increasingly more time planning for and supervising paraprofessionals and less time teaching, leading to the types of role ambiguity or role conflict described in research on working conditions (Brunsting et al., 2014).

Interestingly, special education paraprofessional absences were not significantly correlated with many of the same variables that were correlated with special educator absences. A logical reason may be that paraprofessionals were absent over 80% of their available sick/personal days, compared to less than 61% for special educators. Special education paraprofessionals had fewer days available for leave, so their average absences were closer to their total available leave. This could have been a constraint, limiting the influence of variables that were found correlated with average special educator absences. This means that special education paraprofessionals were absent at high rates regardless of school service delivery variables. While many plausible reasons exist for special education paraprofessional absences (e.g., cold, flu, burnout, lower pay and benefits), examining them are matters for future research.

Limitations

In considering the findings, the reader is encouraged to remain cognizant of the study’s limitations. First, data were collected from a modestly sized, purposeful, convenience sample of schools in Vermont; therefore, generalization to other contexts should not be inferred. A larger, more representative, and diverse sample of schools would permit inclusion of additional school demographic variables. Second, data were provided exclusively at the school level to reduce the burden on schools participating in an exploratory study. As a result, we could not examine individual differences in absences among special education personnel or their perspectives on working conditions, burnout, self-efficacy, or other related factors. Third, all data were cross-sectional, so only correlational relationships were examined. Fourth, we made the decision to forego collecting data on general education teacher absences, which would have allowed for comparisons between different types of teachers, because these data did not directly address our research questions and we were concerned they would add an undue burden on schools that might lower the participation rate. Nevertheless, this pilot examination of special education personnel absences in inclusion-oriented schools has value in illuminating important issues for schools seeking to improve their inclusive service delivery.

Implications and Conclusions

As a pilot investigation, findings suggest that there are notable relationships between special educator absences and a series of school-level predictors; its design does not tell us why these patterns exist. Future studies with more representative samples and longitudinal data collection are needed to estimate causal relationships among these factors and others not examined in this study (e.g., working conditions, burnout, and reasons for absences).

While absences have obvious financial implications for schools (Miller, 2012), most importantly, special education personnel absences may adversely affect continuity of instruction from highly qualified special education and student outcomes. This study points to some initial steps schools can consider. Schools who declined to participate reported difficulties accessing personnel absence data. This suggests one potential action: Include absence data by personnel in school information management systems. This would allow school leaders to compare absences to policy and practice variables at an individual person, school, and district level.

Another potential action is for schools to focus on malleable service delivery factors. While rates of absence tend to be higher in larger schools (see Table 2), such variables typically are not amenable to change. Similarly, the number of available sick/personal days, while positively correlated with special educator absences, often are subject to collective bargaining agreements, so are not readily available as targets of action for school leaders. While a strong correlation was found between average special educator and average paraprofessional absences, this too does not provide clear direction for school leaders because it is unknown if one causes the other or if they are influenced by a common factor (e.g., working conditions; Garwood et al., 2017).

Research from inclusion-oriented schools has offered two special education personnel utilization variables that are malleable (Giangreco et al., 2013), and they were correlated with special educator absences in this study: (a) improving special educator school density by increasing the number of special educator FTE per total enrollment and (b) improving special services concentration by narrowing the ratio of special educators to special education paraprofessionals. These variables are logically linked, since cost-neutral resource reallocation (e.g., trading-in three paraprofessional positions to hire one special educator) has the impact of improving both special educator school density and special services concentration (Giangreco et al., 2011). These variables have been shown to be correlated with each other and more importantly correlated with special educator self-efficacy ratings (Giangreco et al., 2013).

While many factors (e.g., meaningful curriculum, evidence-based instructional practices, positive behavior supports, and principled leadership) are key components of unified educational systems (Burrello, Sailor, & Kleinhammer-Tramill, 2013), arguably no factors are more essential than the quality and availability of the people who deliver educational services (Billingsley, 2011). Simply put, highly qualified special educators only have opportunities to positively contribute to students’ academic, social, and behavioral outcomes if they are consistently present in schools. Attending to these types of service delivery issues helps position schools to get more out of their curricular and instructional efforts by improving continuity of implementation.

Footnotes

Associate Editor: Margaret King-Sears

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Bettini

E. A.

Crockett

J. B.

Brownell

M. T.

Merrill

K. L.

(2016). Relationships between working conditions and special educators’ instruction. The Journal of Special Education, 50, 178–190. doi:10.1177/0022466916644425

Billingsley

B. S.

(2004). Special education teacher retention and attrition: A critical analysis of the research literature. The Journal of Special Education, 38, 39–55. doi:10.1177/00224669040380010401

Billingsley

B. S.

(2011). Factors influencing special education teacher quality and effectiveness. In Kauffman

J. M.

Hallahan

D. P.

(Eds.), Handbook of special education (pp. 391–405). New York, NY: Routledge.

Bowers

(2004). Stress, teaching and teacher health. Education 3-13, 32, 73–80. doi:10.1080/03004270485200361

Bradley

Green

Leeves

(2007). Worker absence and shirking: Evidence from matched teacher-school data. Labour Economics, 14, 319–334. doi:10.1016/j.labeco.2006.05.002

Bright

K. R.

Boone

S. A.

Gerba

C. P.

(2010). Occurrence of bacteria and viruses on elementary classroom surfaces and the potential role of classroom hygiene in the spread of infectious diseases. The Journal of School Nursing, 26, 33–41. doi:10.1177/1059840509354383

Brunsting

N. C.

Sreckovic

M. A.

Lane

K. L.

(2014). Special education teacher burnout: A synthesis of research from 1979 to 2013. Education & Treatment of Children, 37, 681–711. doi:10.1353/etc.2014.0032

Burrello

L. C.

Sailor

Kleinhammer-Tramill

(2013). Unifying educational systems: Leadership and policy perspectives. New York, NY: Routledge.

Christenson

S. L.

Reschly

A. L.

Wylie

(2012). Handbook of research on student engagement. New York, NY: Springer.

10.

Clotfelter

C. T.

Ladd

H. F.

Vigdor

J. L.

(2009). Are teacher absences worth worrying about in the United States? Education Finance and Policy, 4, 115–149. doi:10.1162/edfp.2009.4.2.115

11.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

12.

Duflo

Hanna

Ryan

S. P.

(2012). Incentives work: Getting teachers to come to school. The American Economic Review, 102, 1241–1278. doi:10.1257/aer.102.4.1241

13.

Ehrenberg

R. G.

Ehrenberg

R. A.

Rees

D. I.

Ehrenberg

E. L.

(1991). School district leave policies, teacher absenteeism, and student achievement. The Journal of Human Resources, 26, 72–105. doi:10.2307/145717

14.

Faul

Erdfelder

Buchner

Lang

A. G.

(2009). Statistical power analyses using G*Power 3.1. Behavior Research Methods, 41, 1149–1160. doi:10.3758/BRM.41.4.1149

15.

Garwood

J. D.

Werts

M. G.

Varghese

Gosey

(2017). Mixed-methods analysis of rural special educators’ role stressors, behavior management, and burnout. Rural Special Education Quarterly, 37, 30–43. doi:10.1177/8756870517745270

16.

Giangreco

M. F.

Broer

S. M.

Suter

J. C.

(2011). Guidelines for selecting alternatives to overreliance on paraprofessionals: Field-testing in inclusion-oriented schools. Remedial and Special Education, 32, 22–38. doi:10.1177/0741932509355951

17.

Giangreco

M. F.

Doyle

M. B.

Suter

J. C.

(2012). Constructively responding to requests for paraprofessionals: We keep asking the wrong questions. Remedial and Special Education, 33, 362–373. doi:10.1177/0741932511413472

18.

Giangreco

M. F.

Hurley

S. M.

Suter

J. C.

(2009). Personnel utilization and general class placement of students with disabilities: Ranges and ratios. Intellectual and Developmental Disabilities, 47, 53–56. doi:10.1352/2009.47:53-56 (Erratum published 2010, Intellectual and Developmental Disabilities, 48(1), ii. doi:10.1352/1934-9556-48.1.ii)

19.

Giangreco

M. F.

Suter

J. C.

(2015). Precarious or purposeful? Proactively building inclusive special education service delivery on solid ground. Inclusion, 3, 112–131. doi:10.1352/2326-6988-3.3.112

20.

Giangreco

M. F.

Suter

J. C.

Hurley

S. M.

(2013). Revisiting personnel utilization in inclusion-oriented schools. The Journal of Special Education, 47, 121–132. doi:10.1177/0022466911419015

21.

Gottfried

M. A.

(2010). Evaluating the relationship between student attendance and achievement in urban elementary and middle schools: An instrumental variables approach. American Educational Research Journal, 47, 434–465. doi:10.3102/0002831209350494

22.

Hattie

(2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. New York, NY: Routledge.

23.

Herrmann

M. A.

Rockoff

J. E.

(2012). Worker absence and productivity: Evidence from teaching. Journal of Labor Economics, 30, 749–782. doi:10.1086/666537

24.

Hitti

(2006, February 16). Top 9 jobs where bacteria thrive. WebMD Health News. Retrieved from http://www.webmd.com/a-to-z-guides/news/20060216/9-jobs-bacteria-thrive

25.

Joseph

Waymack

Zielaski

(2014). Roll call: The importance of teacher attendance. Washington, DC: National Council on Teacher Quality. Retrieved from http://www.nctq.org/dmsView/RollCall_TeacherAttendance

26.

Maslach

Leiter

M. P.

(2005). Stress and burnout: The critical research. In Cooper

C. L.

(Ed.), Handbook of stress medicine and health (pp. 155–168). Lancaster, UK: CRC Press.

27.

Miller

R. T.

(2008). Tales of teacher absence: New research yields patterns that speak to policymakers. Washington, DC: Center for American Progress. Retrieved from https://cdn.americanprogress.org/wp-content/uploads/issues/2008/10/pdf/teacher_absence.pdf

28.

Miller

R. T.

(2012). Teacher absence as a leading indicator of student achievement: New national data offer opportunity to examine cost of teacher absence relative to learning loss. Washington, DC: Center for American Progress. Retrieved from https://www.americanprogress.org/wp-content/uploads/2012/11/TeacherAbsence-6.pdf

29.

Miller

R. T.

Murnane

R. J.

Willett

J. B.

(2008). Do teacher absences impact student achievement? Longitudinal evidence from one urban school district. Educational Evaluation and Policy Analysis, 30, 181–200. doi:10.3102/0162373708318019

30.

Monto

A. S.

Malosh

R. E.

Petrie

J. G.

Thompson

M. G.

Ohmit

S. E.

(2014). Frequency of acute respiratory illnesses and circulation of respiratory viruses in households with children over 3 surveillance seasons. The Journal of Infectious Disease, 210, 1792–1799.

31.

Roby

(2013). Teacher attendance effects on student achievement: Research study of Ohio schools. Education, 134, 201–206.

32.

Russ

Chiang

Rylance

B. J.

Bongers

(2001). Caseload in special education: An integration of research findings. Exceptional Children, 67, 161–172. doi:10.1177/001440290106700202

33.

Simon

N. S.

Johnson

S. M.

(2015). Teacher turnover in high-poverty schools: What we know and can do. Teachers College Record, 117(3), 1–36.

34.

Suter

J. C.

Giangreco

M. F.

(2009). Numbers that count: Exploring special education and paraprofessional service delivery in inclusion-oriented schools. The Journal of Special Education, 43, 81–93. doi:10.1177/0022466907313353

35.

Tabachnick

B. G.

Fidell

L. S.

(2013). Using multivariate statistics (6th ed.). Boston, MA: Pearson.

36.

U.S. Department of Education. (2016). 2013-2014 civil rights data collection: A first look. Washington, DC: Author. Retrieved from http://www2.ed.gov/about/offices/list/ocr/docs/2013-14-first-look.pdf

37.

U.S. Department of Education. (2017). Teacher shortage areas nationwide listing 1990–1991 through 2017–2018. Washington, DC: Author. Retrieved from https://www2.ed.gov/about/offices/list/ope/pol/bteachershortageareasreport201718.pdf

38.

U.S. Department of Education. (2018a). Number and percent of students ages 6 through 21 served under IDEA, Part B, by educational environment and state: 2017-18 [Data file]. Washington, DC: Author. Retrieved from https://www2.ed.gov/programs/osepidea/618-data/static-tables/index.html

39.

U.S. Department of Education. (2018b). Public elementary/secondary school universe survey, 2010-11, 2011-12, 2012-13 [Data file]. Washington, DC: Author. Retrieved from https://nces.ed.gov/ccd/pubschuniv.asp

40.

VanVoorhis

C. W.

Morgan

B. L.

(2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology, 3, 43–50. doi:10.20982/tqmp.03.2.p043

41.

Winkler

A. M.

Scull

Zeehandelaar

(2012). How strong are U.S. teacher unions? A state-by-state comparison. Retrieved from https://edexcellence.net/publications/how-strong-are-us-teacher-unions.html

42.

Woods

Montagno

(1997). Determining the negative effect of teacher attendance on student achievement. Education, 118, 307–316.