Skip to main content

Differential Peer Effects, Student Achievement, and Student Absenteeism: Evidence From a Large-Scale Randomized Experiment

Abstract

Using data from a well-executed randomized experiment, I examine the effects of gender composition and peer achievement on high school students’ outcomes in disadvantaged neighborhoods. Results show that having a higher proportion of female peers in the classroom improves girls’ math test scores only in less-advanced courses. For male students, the estimated gender peer effects are positive but less precisely estimated. I also find no effect of average classroom achievement on female math test scores. Males, on the other hand, seem to benefit from a higher-achieving classroom. I propose mechanisms relating to lower gender stereotype influences and gender-specific attitudes toward competition as potential explanations for peer effects findings. Finally, having a higher proportion of female students in the classroom decreases student absenteeism among male students but has no impact on female attendance.

This is a preview of subscription content, access via your institution.

Notes

  1. The number of public schools providing same-sex classrooms increased to approximately 400 schools in 2011 from 12 schools in 2002. There were also more than 110 same-sex schools nationwide in 2011 (Chandler 2011).

  2. A large literature has also examined the effect of peers on student outcomes in college (see, e.g., Carrell et al. 2009; Foster 2006; Lyle 2007; Sacerdote 2001; Stinebrickner and Stinebrickner 2006; Zimmerman 2003). Moreover, recent studies have examined peer effects in labor markets (see, e,g,., Arcidiacono and Nicholson 2005; Black et al. 2013; Bifulco et al. 2011, 2014; Mas and Moretti 2009).

  3. A few studies have also examined the impact of single-sex schooling on student outcomes (see, e.g., Doris et al. 2013; Jackson 2012; Park et al. 2013).

  4. The other major identification issue—referred to as the endogeneity or the reflection problem (Manski 1993; Moffitt 2001; Sacerdote 2001)—occurs because it is often difficult to separate the effect that the peer group has on the student from the effect the student has on the peer group. Suggesting a regression of own achievement on contemporaneous average peer achievement is problematic because these outcomes are jointly determined, and peer achievement is likely to be endogenous to the model. Apart from this threat to identification, peers usually share an environment and thus are potentially subject to similar shocks. Ignoring these correlated effects may also contaminate the estimated peer effects.

  5. Peer aggregation at the classroom level has been limited. See, for example, Antecol et al. (2016), Betts and Zau (2004), Burke and Sass (2013), and Hoxby and Weingarth (2005) for notable exceptions.

  6. The mission of the programs is not limited to secondary schools. Both TFA and TNTP recruit teachers to serve in primary schools in disadvantaged communities.

  7. The control teachers could have entered teaching through either a traditional route or a less-selective alternative certification program.

  8. Because the data are confidential, the sample sizes are rounded to the nearest multiple of 10.

  9. There are 320 teachers in the data (70 TFA, 80 TF, and 170 control group teachers), indicating that some teachers in the study taught in more than one block in different periods during the school day and in both survey years.

  10. The high school study sample includes 110 classroom matches: 120 TFA/TF and 130 control group classrooms from a total of 40 schools. There are also 160 teachers (20 TFA, 50 TF, and 90 control group teachers).

  11. Baseline test scores for all students in the same classroom match come from the same grade level. To ensure this, MPR administrators calculated the tenth percentile of the current grade level within the math classroom matches, identified the highest grade level that was less than the tenth percentile grade level and was a grade level in which end-of-year state assessments were administered, and then used that particular grade level to obtain the baseline test scores.

  12. I also examine whether the baseline characteristics of the attrites are correlated with beginning-of-year classroom measures (e.g., average classroom math achievement) but do not find evidence for any correlation. These results are available upon request.

  13. The average proportion of the compliers in the end-of-year classrooms is approximately 85 %.

  14. Information on student and teacher characteristics is not available for all student observations. I use dummy variables to control for the missing values in student and teacher characteristics.

  15. More than 300 (330) late enrollees had nonmissing baseline math test scores.

  16. An increase in the number of tests increase the likelihood of falsely rejecting the null hypothesis, the so-called multiplicity problem (Anderson 2008). Specifically, of 19 hypotheses, the probability of falsely rejecting at least one of the 19 null hypotheses at the 10 % level is 1 – 0.919 = 0.865. Therefore, rejection of one or two hypotheses among many does not necessarily pose a threat to randomization.

  17. Only 47 noncomplier students switched to a different block. For ease of notation, I use λ b in the equations throughout this article, but I control for the end-of-year block fixed effects in the estimations of peer effects.

  18. As an alternative, I use the grade-level baseline achievement distribution to specify the achievement terciles. The results from this exercise are virtually identical to those presented herein and are available upon request.

  19. The mean of within-block difference in the proportion of female students across classroom matches is .1, with a standard deviation of .09. This difference also ranges between 0 and .55.

  20. I also examine the variance decomposition in the average math achievement of the classrooms. The within-block variation accounts for approximately 4 % of the total variation in the average math achievement.

  21. The coefficient estimates on the instrument from the first stage are 0.718 (SE = 0.075) and 0.750 (SE = 0.073) for female and male subsamples, respectively, and the first-stage F-statistics are 89.92 and 102.96, respectively.

  22. I interpret the coefficient magnitudes using the NWEA nationwide normative sample.

  23. I also examine the effects of peer achievement by excluding classroom gender composition measures. The coefficient estimates on peer achievement effects remain intact. I also add interaction terms between average peer math achievement and gender composition in the classroom to look at any interactive impact. These additional terms are all statistically indifferent from zero.

  24. Using the Angrist-Pischke first-stage statistic, I reject the null hypothesis of weak identification for all five endogenous variables.

  25. Compared with IV regressions, OLS estimates reported in columns 1 and 3 of Table 8 are less significant and are smaller in magnitude.

  26. Apart from the main focus of this study, I examine the effects of student-teacher gender match on math achievement in high schools and fail to find any significant effect of gender match on either female or male students’ achievement.

  27. Following McCrary and Royer (2011), power calculations for a two-sided test are based on the following inequality:

    \( \frac{N- n}{N}>\frac{1.96^2}{\left({\hat{\uptheta}}_N-{\uptheta}_0\right)} S\hat{E}-1, \)

    where θ0 is the point hypothesis to be tested (e.g., 0 in this case), \( {\hat{\uptheta}}_N \) is the point estimate that I expect to obtain in the large sample size N (e.g., the actual point estimates from the randomized sample n), and \( S\hat{E} \) is the estimated standard error using this sample.

  28. I also estimate gender composition and average effects of peer achievement for the subsample of students with nonmissing absentee rates (1,250 females and 1,010 males). The findings for peer effects are very similar, and these additional results are available upon request.

  29. I also estimate a model similar to Eq. (6) where I construct achievement indicators denoting student’s own math baseline achievement at each decile of the course-level baseline achievement distribution. Viewing the complete set of results, it appears that male students in the first five deciles of the baseline achievement distribution seem to benefit the most from an improvement in average math achievement.

  30. I also experiment with the same analysis using fraction of gender-specific peers in the classroom. The results reinforce findings from Table 7 that it is the average of the entire classroom—not just same-sex peers’ achievement—that matters for male students.

  31. The sample weights reflect the probability of assignment of a student to a TFA/TF classroom.

  32. In addition to these robustness checks, I first estimate the gender peer effects separately for TFA and TNTP subsamples. Next, I interact student and teacher controls with indicators for student’s position in the math baseline achievement distribution. The coefficient estimates are similar to those reported throughout this article.

  33. Comparing the mean, median, as well as the range of the proportion of female students in the classrooms, I do not observe any discernible patterns across math subjects.

  34. Whitmore (2005) also reported the pooled sample results from a specification in which the share of female peers and average achievement of the classroom were simultaneously controlled for. The estimated effect of the share of female students falls from 2 percentile points to 1.3 when achievement of the classroom is included to the gender peer effects specification.

  35. Psychology literature offers ample evidence regarding the relationship between peer influences and age. Conformity to peer pressure follows an inverted U-shaped age pattern, with peer effects peaking during mid-adolescence (see, e.g., Blakemore and Choudhury 2006; Brown et al. 1986).

References

  • Anderson, M. D. (2012). In school and out of trouble? The minimum dropout age and juvenile crime. Review of Economics and Statistics, 96, 318–331.

  • Anderson, M. L. (2008). Multiple inference and gender differences in the effects of early intervention: A reevaluation of the Abecedarian, Perry Preschool, and Early Training projects. Journal of the American Statistical Association, 103, 1481–1495.

    Article  Google Scholar 

  • Antecol, H., Eren, O., & Ozbeklik, S. (2016). Peer effects in disadvantaged primary schools: Evidence from a randomized experiment. Journal of Human Resources, 51, 95–132.

    Article  Google Scholar 

  • Arcidiacono, P., & Nicholson, S. (2005). Peer effects in medical school. Journal of Public Economics, 89, 327–350.

    Article  Google Scholar 

  • Berthelon, M. E., & Kruger, D. I. (2011). Risky behavior among youth: Incapacitation effects of school on adolescent motherhood and crime in Chile. Journal of Public Economics, 95, 41–53.

    Article  Google Scholar 

  • Betts, J. R., & Zau, A. (2004). Peer groups and academic achievement: Panel evidence from administrative data. Unpublished manuscript, University of California, San Diego, CA, and Public Policy Institute of California, San Francisco, CA.

  • Bifulco, R., Fletcher, J. M., Oh, S. J., & Ross, S. L. (2014). Do high school peers have persistent effects on college attainment and other life outcomes? Labour Economics, 29, 83–90.

    Article  Google Scholar 

  • Bifulco, R., Fletcher, J. M., & Ross, S. L. (2011). The effect of classmate characteristics on post-secondary outcomes: Evidence from the Add Health. American Economic Journal: Economic Policy, 3(1), 25–53.

  • Black, S. E., Devereux, P. J., & Salvanes, K. G. (2013). Under pressure? The effect of peers on outcomes of young adults. Journal of Labor Economics, 31, 119–153.

  • Blakemore, S. J., & Choudhury, S. (2006). Development of the adolescent brain: Implications for executive function and social cognition. Journal of Child Psychology and Psychiatry, 47, 296–312.

    Article  Google Scholar 

  • Brown, B. B., Clasen, D. R., & Eicher, S. A. (1986). Perceptions of peer pressure, peer conformity dispositions, and self-reported behavior among adolescents. Developmental Psychology, 22, 521–530.

    Article  Google Scholar 

  • Burke, M. A., & Sass, T. R. (2013). Classroom peer effects and student achievement. Journal of Labor Economics, 31, 51–82.

    Article  Google Scholar 

  • Carrell, S. E., Fullerton, R. L., & West, J. E. (2009). Does your cohort matter? Measuring peer effects in college achievement. Journal of Labor Economics, 27, 439–464.

    Article  Google Scholar 

  • Carell, S. E., Sacerdote, B. I., & West, J. E. (2013). From natural variation to optimal policy? The importance of endogenous peer group formation. Econometrica, 81, 855–882.

  • Chandler, M. A. (2011, September 22). Single-sex education may do more harm than good. The Washington Post. Retrieved from https://www.washingtonpost.com/local/education/study-single-sex-education-may-do-more-harm-than-good/2011/09/22/gIQABAQOoK_story.html?utm_term=.04abcd6321da

  • Clark, M. A., Chiang, H. S., Silva, T., McConnell, S., Sonnenfeld, K., & Erbe, A. (2013). The effectiveness of secondary math teachers from Teach for America and the Teaching Fellows programs (NCEE Report No. 2013–4015). Washington, DC: U.S. Department of Education.

  • Doris, A., O’Neill, D., & Sweetman, O. (2013). Gender, single-sex schooling and maths achievement. Economics of Education Review, 35, 104–119.

    Article  Google Scholar 

  • Falk, A., & Ichino, A. (2006). Clear evidence on peer effects. Journal of Labor Economics, 24, 39–57.

    Article  Google Scholar 

  • Foster, G. (2006). It’s not your peers, and it’s not your friends: Some progress toward understanding the educational peer effect mechanism. Journal of Public Economics, 90, 1455–1475.

    Article  Google Scholar 

  • Gneezy, U., Niederle, M., & Rustichini, A. (2003). Performance in competitive environments: Gender differences. Quarterly Journal of Economics, 118, 1049–1074.

    Article  Google Scholar 

  • Graham, B. S. (2008). Identifying social interactions through conditional variance restrictions. Econometrica, 76, 643–660.

    Article  Google Scholar 

  • Guryan, J., Kroft, K., & Notowidigdo, M. J. (2009). Peer effects in the workplace: Evidence from random groupings in professional golf tournaments. American Economic Journal: Applied Economics, 1(4), 34–68.

    Google Scholar 

  • Hanushek, E. A. (2011). The economic value of higher teacher quality. Economics of Education Review, 30, 466–479.

    Article  Google Scholar 

  • Hanushek, E. A., Kain, J. F., Markman, J. M., & Rivkin, S. G. (2003). Does peer ability affect student achievement? Journal of Applied Econometrics, 18, 527–544.

    Article  Google Scholar 

  • Hoxby, C. M. (2000). Peer effects in the classroom: Learning from gender and race variation (NBER Working Paper No. 7867). Cambridge, MA: National Bureau of Economic Research.

  • Hoxby, C. M., & Weingarth, G. (2005). Taking race out of the equation: School reassignment and the structure of peer effects. Unpublished manuscript. Retrieved from https://www.pausd.org/sites/default/files/pdf-faqs/attachments/TakingRaceOutOfTheEquation.pdf

  • Imberman, S. A., Kugler, A. D., & Sacerdote, B. I. (2012). Katrina’s children: Evidence on the structure of peer effects from hurricane evacuees. American Economic Review, 102, 2048–2082.

    Article  Google Scholar 

  • Jackson, K. C. (2012). Single-sex schools, student achievement, and course selection: Evidence from rule-based student assignments in Trinidad and Tobago. Journal of Public Economics, 96, 173–187.

    Article  Google Scholar 

  • Jacob, B. A., & Lefgren, L. (2003). Are idle hands the devil’s workshop? Incapacitation, concentration, and juvenile crime. American Economic Review, 93, 1560–1577.

    Article  Google Scholar 

  • Langdon, D., McKittrick, G., Beede, D., Khan, B., & Doms, M. (2011). STEM: Good jobs now and for the future (ESA Issue Brief No. 03–11). Washington, DC: U.S. Department of Commerce.

  • Lavy, V., & Schlosser, A. (2011). Mechanisms and impacts of gender peer effects at school. American Economic Journal: Applied Economics, 3(2), 1–33.

    Google Scholar 

  • Lu, F., & Anderson, M. L. (2015). Peer effects in microenvironments: The benefits of homogeneous classroom groups. Journal of Labor Economics, 33, 91–122.

  • Luallen, J. (2006). School’s out… forever: A study of juvenile crime, at-risk youths and teacher strikes. Journal of Urban Economics, 59, 75–103.

    Article  Google Scholar 

  • Lyle, D. S. (2007). Estimating and interpreting peer and role model effects from randomly assigned social groups at West Point. Review of Economics and Statistics, 89, 289–299.

    Article  Google Scholar 

  • Manski, C. F. (1993). Identification of endogenous social effects: The reflection problem. Review of Economic Studies, 60, 531–542.

    Article  Google Scholar 

  • Mas, A., & Moretti, E. (2009). Peers at work. American Economic Review, 99, 112–145.

    Article  Google Scholar 

  • McCrary, J., & Royer, H. (2011). The effect of female education on fertility and infant health: Evidence from school entry policies using exact date of birth. American Economic Review, 101, 158–195.

    Article  Google Scholar 

  • Moffitt, R. A. (2001). Policy interventions, low-level equilibria, and social interactions. In S. N. Durlauf & H. P. Young (Eds.), Social dynamics (pp. 45–82). Cambridge, MA: MIT Press.

    Google Scholar 

  • Niederle, M., & Vesterlund, L. (2010). Explaining the gender gap in math test scores: The role of competition. Journal of Economic Perspectives, 24(2), 129–144.

    Article  Google Scholar 

  • Park, H., Behrman, J. R., & Choi, J. (2013). Causal effects of single-sex schools on college entrance exams and college attendance: Random assignment in Seoul high schools. Demography, 50, 447–469.

    Article  Google Scholar 

  • Pope, D. G., & Sydnor, J. R. (2010). Geographic variation in the gender differences in test scores. Journal of Economic Perspectives, 24(2), 95–108.

    Article  Google Scholar 

  • Sacerdote, B. I. (2001). Peer effects with random assignment: Results for Dartmouth roommates. Quarterly Journal of Economics, 116, 681–704.

    Article  Google Scholar 

  • Sacerdote, B. I. (2011). Peer effects in education: How might they work, how big are they and how much do we know thus far? In E. Hanushek, S. Machin, & L. Woessmann (Eds.), Handbook of the economics of education (Vol. 3, pp. 249–277). Amsterdam, The Netherlands: Elsevier.

    Google Scholar 

  • Schemo, D. J. (2006, October 25). Federal rules back single-sex public education. The New York Times. Retrieved from http://www.nytimes.com/2006/10/25/education/25gender.html

  • Sojourner, A. (2013). Identification of peer effects with missing peer data: Evidence from Project STAR. Economic Journal, 123, 574–605.

    Article  Google Scholar 

  • Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women’s math performance. Journal of Experimental Social Psychology, 35, 4–28.

    Article  Google Scholar 

  • Stinebrickner, R., & Stinebrickner, T. R. (2006). What can be learned about peer effects using college roommates? Evidence from new survey data and students from disadvantaged backgrounds. Journal of Public Economics, 90, 1435–1454.

    Article  Google Scholar 

  • Sullivan, A., Joshi, H., & Leonard, D. (2012). Single-sex and coeducational secondary schooling: What are the social and family outcomes in the short and longer term? Longitudinal and Life Course Studies, 3, 137–156.

    Google Scholar 

  • Vigdor, J. L., & Nechyba, T. (2007). Peer effects in North Carolina public schools. In L. Woessmann & P. E. Peterson (Eds.), Schools and the equal opportunity problem (pp. 73–102). Cambridge, MA: MIT Press.

    Google Scholar 

  • Whitmore, D. (2005). Resource and peer impacts on girls’ academic achievement: Evidence from a randomized experiment. American Economic Review: Papers and Proceedings, 95, 199–203.

    Article  Google Scholar 

  • Zimmerman, D. J. (2003). Peer effects in academic outcomes: Evidence from a natural experiment. Review of Economics and Statistics, 85, 9–23.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Ozkan Eren.

Electronic supplementary material

Online Resource 1

(PDF 123 kb)

Online Resource 2

(PDF 25 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Eren, O. Differential Peer Effects, Student Achievement, and Student Absenteeism: Evidence From a Large-Scale Randomized Experiment. Demography 54, 745–773 (2017). https://doi.org/10.1007/s13524-017-0552-8

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s13524-017-0552-8

Keywords

  • Peer effects
  • Instrumental variables
  • Randomized experiment
  • Teach for America
  • Power calculations