As schools in many regions of the world struggle to employ highly qualified and effective teachers in every classroom, governments are experimenting with different strategies to increase the supply of new teachers. One of the most popular strategies being attempted is to allow people to become teachers with either no training or with training that is faster and cheaper than degree-based teacher preparation programs (TPPs; Author, 2022). These new, faster and cheaper TPPs are non-degree-based programs. Like teacher candidates enrolled in the original French École Normale Supérieure (Normal School), established in 1794, candidates in these new TPPs are only prepared to teach and do not earn an academic degree.  These new TPPs are non-degree programs just like the majority of TPPs that existed around the world from the 1700s through the 1960s. Despite this long history of TPPs being non-degree-based, these new TPPs are called “alternative.” Some state governments in the United States, like Texas, have been experimenting with these alternative programs for several decades, and as a result, Texas now prepares over 50% of the alternatively prepared teachers in America. There is evidence that these alternative programs are effective at preparing large numbers of teachers and at preparing a greater diversity of new teachers (more male teachers, more teachers of color; Author, 2019). Unfortunately, these alternatively prepared teachers are also more likely to leave the profession soon after entering the classroom compared to traditionally prepared teachers (Author, 2019).

One of the primary differences between traditional, degree-based TPPs and alternative, non-degree-based TPPs is the type of clinical teaching experience they provide (Darling-Hammond, Chung, & Frelow, 2002). The vast majority of alternatively prepared teachers complete an internship during which they serve as the official teacher-of-record employed by the school, and they receive a full teacher’s salary. By contrast, the vast majority of traditionally prepared teachers complete an unpaid student teaching clinical experience under the mentoring and supervision of a teacher-of-record employed by the school, and the teacher candidate must successfully complete their TPP before they themselves can become an official teacher-of-record.  

The impacts of these two types of clinical teaching experiences on teacher self-efficacy has been examined and the results are generally mixed with a few studies showing traditionally prepared teachers have higher self-efficacy than alternatively prepared teachers (Darling-Hammond et al., 2002; Zientek, 2007), but the majority shows no differences (e.g., Fox & Peters, 2013; Griffin, 2022; Jackson & Miller, 2020). Given that teaching self-efficacy is positively correlated with student learning and that these results are mixed, we cannot draw a firm conclusion about the impacts of clinical experience types on student learning. Direct comparisons of these different clinical teaching methods on secondary students’ academic growth have not been examined. Our primary research question is, are first-time teachers-of-record equally effective at impacting secondary student academic growth in English/Reading and Mathematics regardless of the type of clinical teaching they experienced? Our secondary question is, are new teachers-of-record with no teaching preparation or with preparation but in a different subject area as effective as the new teachers-of-record who engaged in either clinical teaching experience?

If governments are moving to allow alternative TPPs and these programs result in similar student academic growth as traditional TPPs while also being faster and cheaper, then governments are prudent to move towards allowing alternative TPPs. If, however, the alternatively prepared teachers are harming the academic growth of students, then caution is certainly warranted, and changes to the alternative TPP curricula should be strongly encouraged or mandated. Similarly, if unprepared people can become teachers who are equally effective, then why require any teacher preparation at all?

To answer our research questions, we used the Texas state longitudinal data system (TLDS), which contains almost three decades of detailed student and teacher data for all 5.4 million primary and secondary students annually enrolled in Texas public schools. The TLDS also contains state assessment scores for standardized English/Reading and Mathematics exams. Since 2011, the TLDS also includes data to link students to teachers to specific classes. The TLDS allows us to know which teachers taught which students and in which courses and then how the student performed on the state exams during the year of instruction and in prior school years. Using the TLDS, we built two custom datasets that covered the 2011-12 through 2018-19 school years. One dataset included 12.9 million students enrolled in English/Reading classes in Grades 7, 8, 9, and 10, and one dataset included 9 million students in Mathematics in Grades 7, 8, and 9. Only students who were taught by new teachers-of-record were included in the final datasets; this resulted in approximately 742,000 English/Reading students and 478,000 Mathematics students. We assessed the impacts of these new teachers of record on their students’ academic growth in English/Reading and Mathematics by estimating three-level hierarchical linear models (HLM) for each grade level and/or subject. Students were at level 1, teachers at level 2, and schools at level 3, and the data were strictly nested so that each student existed under only one teacher, and each teacher existed under only one school. The dependent variables were the normalized scaled score on the corresponding state assessment. The student-level predictors included the normalized state assessment score during the prior school year in the same subject, student gender, student race/ethnicity, English-language learner status, economic status, and special education status. The teacher-level predictor was the type of clinical teaching experience (intern versus student teacher versus no preparation). School- predictors include locale (e.g., urban suburban, rural), percentage of students who are economically disadvantaged, and the percentage of students who were white. The non-binary variables were normalized (z-score transformed) using the state average for that variable. All of these variables have been shown in our prior research to be correlated with the assessment outcome (e.g., Author, 2022). The intraclass correlation shows that a large percentage of variance is accounted for at the teacher and school levels, thereby indicating that HLM is warranted (Snijders & Bosker, 2012).

Our primary research question is: Are first-time teachers-of-record equally effective at impacting secondary students’ academic growth in English/Reading and Mathematics regardless of whether they completed a student teaching clinical experience or they were engaged in an internship-based clinical experience? The answer is: Interns are significantly less effective as new teachers-of-record compared to those teachers-of-record who successfully completed a student teaching clinical experience. This pattern exists for both English/Reading and Mathematics, and across all grade levels (Grades 7 to 10). The magnitude of the negative impact on student learning of teachers-of-record engaged in an internship was up to twice the impact of student poverty. In other words, ensuring that intern teachers-of-record were as well prepared as teachers-of-record who completed student teaching would have up to two times the positive benefit on student learning as eliminating poverty for millions of students in Texas. These results show students learn significantly less when taught by intern teachers-of-record, and strongly suggest the current laws and rules governing alternative TPPs in Texas are insufficient to protect school children. To ensure teacher candidates are better prepared to take on the responsibility of being the official teacher-of-record, the curricula prior to internships need to be revised and strengthened. Our secondary research question is: are unprepared teachers-of-record equally effective at impacting student academic growth as teachers-of-record who were interns or student teachers? The answer has two pieces. First, across all grade levels, the unprepared teachers are significantly less effective than the teachers-of-record who completed student teaching. Second, in some grade levels the unprepared teachers-of-record are significantly less effective than the intern teachers-of-record, and in most cases, the unprepared are equally ineffective. The policy implication of this result is that people who are unprepared to teach should not be hired as teachers-of-record except in very limited, emergency-only situations.

Author. (2019). Author. (2022). Darling-Hammond, L., Chung, R., & Frelow, F. (2002). Variation in teacher preparation: How well do different pathways prepare teachers to teach? Journal of Teacher Education, 53(4), 286–302. Fox, A.G., & Peters, M.L. (2013). First Year Teachers: Certification Program and Assigned Subject on Their Self-Efficacy. Current Issues in Education, 16. Griffin, N. M., "A Comparative Study of Self-Efficacy Between Teachers in Traditional or Alternative Certification Pathways" (2022). Doctoral Dissertations and Projects. 3935. Jackson, N., & Miller, R. (2020). Teacher Candidates’ Sense of Self-Efficacy Toward Classroom Management. Journal of Education, 200(3), 153–163. Snijders, T. A. B., & Bosker, R. J. (2012). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed.). Sage. Zientek L. R. (2007). Preparing high-quality teachers: Views from the classroom. American Educational Research Journal, 44(4), 959–1001.