Recently, while redesigning a course called “Adult Motivation and Conflict Management in Education Settings,” I had a chance to revisit some research on the practice of induction for new teachers. This has made me reflect not only on the importance of the practice, but also on how much more we have to learn in order to slow the huge numbers of new teachers leaving the profession—a trend that drains school budgets and the time and energy of administrators.

When I was a school district superintendent, I believed that there were two important motivators for teachers to improve their practice and to stay in the profession and in their schools: (1) a belief that their instruction is successful in helping students learn well, and (2) a sense that they are supported in achieving success with instruction by their principal and their colleagues. An important part of supporting teachers is offering induction programs to beginning teachers. But the research on the efficacy of induction is not so easily interpreted, and it may be that many induction programs aren’t offering the types of support associated with success.

My perspective has been informed a great deal by two pieces of research conducted by Richard Ingersoll, a University of Pennsylvania professor and a leading researcher on teacher recruitment, retention, and shortages. In a 2004 study published in the American Educational Research Journal, Ingersoll and Thomas M. Smith of Vanderbilt University reported that induction and turnover were associated. That is, teachers who participated in induction programs were more likely to stay in the profession and in their schools. Elements of an induction program that were most associated with beginning teachers remaining in the profession included having an assigned mentor who was teaching the same grade and/or subject, participating in common planning time with colleagues teaching the same grade and/or subject, and participating with colleagues in collaborative meetings regarding instruction issues.

In a recent 2011 review of the research on the effects of teacher induction, Ingersoll and Michael Strong of the University of California, Santa Cruz, documented a doubling of school district induction programs, but also, curiously enough, a climbing attrition rate. Their review began with a search that discovered about 500 studies related to induction, published from the mid-1980s to 2011. However they excluded all but the 15 studies that were empirical, addressed effects on at least one of three outcomes—teacher commitment and retention, teacher instructional practices, and student achievement—and provided enough information to check results. According to their review, the number of first-year teachers in the country has burgeoned from 50,000 in 1987–88 to 200,000 in 2007–08. Additionally, they report that the modal (most frequent) years of experience among all teachers plummeted from 15 years in the late 80s to less than one year in 2008. Estimates vary regarding the attrition rate, but it appears that between about 45 percent and 50 percent of teachers leave the profession within five years, and that attrition among first-year teachers is growing. In recognition of the educational and financial costs of attrition, induction has also grown, as indicated by reports of involvement in induction programs from about 40 percent of first-year teachers in 1990 to nearly 80 percent by 2008. My state, Virginia, and 21 others, moreover, began requiring school districts to operate induction programs in 2008.

So what does all this mean? It is unclear to me why this increase in induction has been accompanied by an uptick in the attrition rate for first-year teachers. I would speculate, but certainly don't know for sure, that the increase in the proportion of new teachers in the teaching force accounts for the increase in attrition rates for two reasons. First of all, there are more first-year teachers and, perhaps because of the increased number needed, schools and districts have become less selective in hiring, resulting in a larger proportion of novice teachers who are experiencing difficulties. Secondly, the greater number of new teachers places a heavier and perhaps less manageable burden on schools and districts, making it more difficult to conduct high-quality induction initiatives and to retain first-year teachers and help improve their performance.

The answer to the question of how to stem the turnover of new teachers appears to lie in the kind of induction provided. The data are less rich in that regard. That is, it is less clear which elements of induction are most strongly related to retention. Ingersoll’s data suggest that some combination of principal support, collegial support, and mentoring by a teacher in the same grade and/or subject are most important individually and even more powerful together.

Although Ingersoll and Strong reported considerable limitations for all of the studies they reviewed, they nevertheless concluded that, in the districts where these 15 studies took place, “induction for beginning teachers and teacher mentoring programs in particular have a positive impact,” and that “almost all of the studies we reviewed showed that beginning teachers who participated in some kind of induction had higher satisfaction, commitment, or retention.” These teachers also performed better in the classroom and their students had greater gains on academic achievement tests.

These tentative results for those engaged in induction efforts are nevertheless encouraging, because they at least identify a few interventions that appear to reduce the probability of new teachers fleeing their schools or the profession. In addition, Ingersoll and Strong included, along with these results, a set of useful questions that researchers and school leaders should consider as they design, evaluate, and study their teacher induction efforts. Some of these questions include

• What content should be featured in induction programs (e.g., subject matter, instructional strategies, and/or classroom management)?
• Which content has a greater impact on desired results (e.g., student performance on state criterion-referenced tests and engaging students in higher-order thought)?
• Should teachers be involved in induction efforts for one, two, or three years (the current range) and which components of induction should be featured for how long?
• What are the relative costs and benefits of induction for the school and for student achievement?
• What variation in contexts (e.g., suburban, rural, urban, high poverty, low poverty, well-prepared new teachers, less well-prepared new teachers) allows some features of induction programs to be more effective than others?

Taken together, these findings and questions lead me to recommend that those who plan, conduct and evaluate induction programs (1) continue them; (2) include in the first year of induction at least those components found to reduce the likelihood of first-year teachers leaving their school or profession (i.e., assign mentors who teach the same grade and/or subject as the first year teacher, make provisions for common planning time for the novice teachers and their same subject and/or grade colleagues, and arrange collaborative sessions on instructional issues for new teachers with their more experienced colleagues); (3) assess the usefulness of a second and/or third year of induction; and (4) study the impact of the induction program on student achievement.

Finally, I continue to believe that these findings about induction (coupled with the literature on the importance of a sense of instructional efficacy in teacher retention—a topic for a future column) support the notion that the most the vital factors in retaining teachers are teachers’ beliefs that they are having a salutary effect on student learning and that their principal and colleagues are supporting them in that effort.

Robert G. Smith is a former superintendent of the Arlington (VA) Public Schools and an associate professor in the College of Education and Human Development at George Mason University.