We also examined the Year 5 graduation outcomes for persisters and dropouts (see Figure 2). By the end of the fifth year, over one-third (38%) of the 4^th year non-graduates had earned a regular high school diploma from a Maryland public high school and 4% had earned a diploma via GED. However, more than half (58%) of the 4^th year non-graduates had not earned any high school credential by the end of Year 5. Nearly half (47%) of persisters earned a high school diploma compared to fewer than 1 in 20 (4%) of dropouts. Dropouts did however earn Maryland diplomas via GEDs at a substantially higher rate (15%) than persisters (2%). Critically, 4 out of 5 (81%) dropouts and just over half (52%) of persisters did not earn a credential by the end of Year 5.

Next, we examined patterns of enrollment and postsecondary participation as well as early workforce participation in Year 6 by Year 5 high school completion status. Just under one-fifth of late graduates (18%) and diploma via GED earners (21%) were enrolled in a postsecondary institution in Year 6 compared to 3% of students who had not yet earned a high school credential (Figure 3). During the sixth year, almost half of diploma via GED earners (44%) and two-fifths of late graduates (38%) worked three or more financial quarters in Maryland’s formal, non-federal economy, compared to less than a quarter of students with no earned high school credential (23%). About half of the late graduates (47%) and diploma via GED earners (45%) did not enroll in postsecondary or work at least 3 quarters during Year 6 compared to three-fourths (75%) of students with no earned high school credential.

Finally, multilevel modeling was used to examine the predictors of dropout and persisting among fourth-year high school non-completers. Figure 4 presents an illustration of the odds ratios and corresponding 95% confidence intervals for the student- and school-level factors included in the final model. For each predictor variable, the dot represents the odds ratio (i.e. the change in likelihood of persisting versus dropping out) and the solid line represents the upper and lower confidence intervals (i.e. range of plausible values for the odds ratio). If the confidence interval crosses one—represented by the perpendicular dotted line—the predictor variable was not significantly related to the odds of persisting versus dropping out (p> .05). When the confidence interval does not cross one, the predictor variable was significantly related to the odds of persisting versus dropping out (p< .05). When the odds ratio is larger than one, that variable is significantly related to higher odds of persisting. When the odds ratio is less than one, the variable was significantly related to reduced odds of persisting.

After controlling for other variables in the model, multiple student-level variables that were related to the odds of persisting or dropping out at the end of Year 4 were identified:

• Black students were more likely than their White peers to be persisters than to have dropped out by the end of Year 4.
• Special education students were more likely to persist than to drop out.
• Students who passed their English HSA by the beginning of Year 4 were more likely to persist than to drop out.
• Students who had dropped out early (years 1-3), but reenrolled by Year 4 were less likely to be persisters (i.e. more likely to drop out) in Year 4.
• Students who were behind in credits by one year (or more) were less likely to persist than to have dropped out by the end of Year 4.
• Students who had a history of chronic absenteeism were less likely to persist than to have dropped out by the end of Year 4.

In addition, multiple school-level variables were related to increased odds of persisting or dropping out at the end of Year 4, including attending a Vocational-Technical or a Combined (i.e. 6-12^th grade) high school (compared to those who attended a traditional high school). In addition, students who attended larger schools had higher odds of persisting versus dropout. 

Policy Implications

There is no current policy in the State of Maryland that specifically identifies persisting students or considers their educational or workforce needs. Our findings, highlighting the unique characteristics and post-secondary outcomes of persisting students, as well as the relatively large size of this group, underscore the need for the State to consider persisting alongside dropout when developing policies aimed at increasing high school completion and promoting positive postsecondary and workforce outcomes. Persisters likely benefit from policies aimed at preventing dropout, but our findings suggest that a shift to invest public resources specifically for persisting students would help to improve graduation rates and subsequent academic and workforce outcomes. Some persisting students may need minimal academic support and coaching to complete the requirements for a diploma in the summer or fall following their fourth year of high school. Others may have more complicated needs as they approach emerging adulthood (e.g., parenting, learning challenges, homelessness) that would require additional resources and intervention (e.g. case management, wraparound services) that may stretch into their fifth or even sixth year of high school. Students who persist this long likely also need social-emotional supports as they become cut off from their former peers who have gone on to college or career.

 Method and Measures

This study first used descriptive statistics to describe the patterns of academic and early workforce participation for four-year non-graduates including identifying patterns in students’ fifth and sixth year academic, postsecondary and early workforce outcomes. Next, multilevel modeling with students nested within high schools^a was used to examine the student- and school-level predictors of dropout versus persisting among fourth-year high school students.

This study included students enrolled in a Maryland public high school as a first-time 9^th grade student for a period of 90 days or more in Year 1 (academic year 2009-10) and attended a Maryland high school in Years 4 (2012-13) or 5 (2013-14). From this cohort of 64,130 students the analysis excluded:

• 12,537 students who were not enrolled in a Maryland public high school in Year 4, including 2,660 students who dropped out in Years 1-3 (early dropouts^b) and did not reenroll in a Maryland public school by the end of Year 4;
• 132 students who were working towards a certificate of completion rather than a regular high school diploma; and
• 46,499 students who graduated in Year 4.

After applying the exclusion criteria, 4,962 students met the final inclusion criteria, constituting a cohort of students who attended a Maryland public high school in Year 4 or 5, and either dropped out or persisted toward a high school diploma. Of those students 20% (n=971) dropped out during Year 4 and 80% were persisters (n=3,991). It is important to note that students identified as “dropping out” could later re-enroll. In addition, the current study used data collected before Maryland raised the age of required school attendance from 16 to 18. Table 1 shows the demographic characteristics for persisters and dropouts.

^a. This study controlled for the third level of nesting (schools within districts).

^b. Early dropouts (students who dropout prior to year 4) do not reengage with high school in large numbers (2%), so this analysis focused on examining and improving our understanding of Year 4 dropouts and persisting students.

Measures

Outcome Variables

Post-Year 4 High School Enrollment indicates that a student enrolled in a Maryland public high school during the summer or academic year immediately following their expected graduation date.

Late graduation indicates whether or not a student completed the requirements for a Maryland high school diploma after their expected graduation date, but before August 1, 2014 (Year 5).

GED indicates whether or not a student completed the requirements for a Maryland diploma via GED before June 2014.

Persister indicates that a student was still enrolled on their expected graduation date, but had not earned a regular high school diploma.

Year 4 Dropout indicates that a student formally withdrew from school without reenrolling in a degree granting program in their 4^th year of high school.

Workforce participation indicates that a student worked 3 or more financial quarters in Year 6. The available data includes students working for employers in Maryland subject to unemployment insurance. The data do not include workforce participation for military service, federal government employment, self-employment, or independent contractors.

Postsecondary enrollment indicates that a student enrolled in a Maryland or out-of-state postsecondary institution during Year 6.

Student Characteristics Used in Multilevel Model Predicting Persisting and Dropout

Student-level characteristics measured in high school were used to predict the odds of persisting and dropping out. The variables examined included gender (male vs. female), Hispanic, English learner, Free and Reduced Priced Meals (FARMs) eligible (i.e. households with incomes at or below 185% of the federal poverty level or eligible for programs including migrant, homeless, and foster care), special education service receipt, and being identified as having experienced homelessness by their school system. Three dummy variables were created for race (Black, White, and Other-race).

Mobility was calculated by assessing whether a student changed schools at least once during the study period. Indicators for whether a student had dropped out of high school in Years 1-3, was chronically absent (i.e. attended < 80% of days in a school year), or had ever been disciplined (i.e. suspended or expelled) were included. To describe student academic performance, a variable describing whether a student was on-track (i.e. had enough credits to be a 12^th grader in Year 4) and if not, how many years behind, in terms of credits, the student was at the beginning of Year 4 were created. Secondly, a variable indicating whether the student had passed the English high school assessment (HSA; an end of course exam that was required for graduation in Maryland) by the beginning of Year 4 was created.

High School Characteristics Used in Multilevel Model Predicting Persisting and Dropout

School type (i.e. Traditional High school, Combined School, Vocational-Technical School, Alternative School, or Charter School). There was a total of 198 schools in the sample, of which 70% were traditional high schools, 7% were combined schools, 7% were vocational schools, 11% were alternative schools, and 6% were charter schools. School size (i.e. the total student-population) was included as well.