Under Water Retention

Just as research into classroom best practices is equivocal, so too is the research related to attainment and retention.  It seems that the more we find out, the less we know.  The less we know, the more difficult a time we have defending what we’re doing, fixing what we perceive is a problem.

We spend a lot of time studying students:  their educational background, their ethnicity, race, socio-economic status (SES), growth mindedness, levels of “engagement.” Or we look at institutional behavior:  program effectiveness, developmental education (see previous post), New Student Seminar-type classes, advising resources and practices.

Yet Wyman found that 65% of retention could be explained by two exogenous variables:  the ratio of institutional academic and support spending per student head count and regional employment per capita.  This actually makes sense—many of our older students are in school to improve their credentials in order to get better jobs.  If the job market is favorable to workers, students have less incentive to stick around.  Similarly, if their need for advising services is frustrated or they find class sizes overwhelming, it is reasonable to imagine them being daunted and bail.  Still, researchers otherwise seem to ignore economic or institutional funding variables in their hunt for THE ANSWER.

If nothing else, it should warn us how far we are from understanding what we’re dealing with.  It appears to be more complex than we can fathom.  And if we don’t understand the cause—or causes—then our responses will also be suspect, even if results are initially positive. 

Some researchers think the main connection is between school and career goals (Luke, et al; Achieving the Dream; NY’s ASAP).  They stress full-time enrollment and counselor support.  Trouble is, a large majority of community college students can’t manage a full-time schedule.  How you come up with THE ANSWER that doesn’t fit the profile of most of your students is possibly at the heart of what is wrong with the reform movement—people supplying solutions that redefine the scope and purpose of community colleges to gain efficiencies.  Do we really want to go in that direction and make community colleges “successful” by merely ridding them of problematic students?  Is that the final solution?

Others seek to test a variety of possible factors associated with retention and failure with the view to constructing programs and practices to enhance the first and minimize the second.  The plethora of factors studied or ignored is such that it is difficult to get any sense of what significance the studies have beyond meeting a department publishing quota.  The factors might include age, gender, ethnicity, race, SES (rarely), course load, remedial placement, receipt of financial aid, tutor use, technology aptitude (Mertes and Hoover: 2012), parent’s college level, high school GPA, college GPA, school rank, ACT score, and a host of psychological characteristics, including amiability, intent to return, school to work connections, personal agency, conscientiousness, “engagement,” and so on. 

There are several problems with the procedures followed to gain the relevant information about these factors.  First, the data are often unreliable or sketchy, either because it is self-reported or incomplete.  High school GPA is unreliable because it is often unavailable and researchers rely on students’ recall.  Furthermore, how representative of a student’s knowledge and academic skills, as well as the soft skills required of successful students, is largely dependent on the rigor imposed by high school, which can vary greatly.  Race is becoming more difficult to sort out all the time, as intermarriage has introduced the complications of multiracial identity (Patten: 2015):  the “one drop of blood” test is headed toward the oblivion it deserves.  While psychological tests may exhibit internal reliability—the results vary little from different dates with the same sample—their ability to describe real qualities outside of behavioral observation is contestable, particularly as the data are self-reported, and consistency does not mitigate self-deception (Goleman: 1985; Brown, 102ff: 2014; Sanders and Katz: 2013).

Some variables make sense, but generate further problems.  “Engagement,” loosely defined, is about how much studying and homework a student does.  It sounds like a no-brainer factor in student success.  A committed student, even if she is not always using the most effective methods to study and do homework, will do better than most students who spend little time on their coursework.  But this isn’t anything teachers don’t already know.  The question is how to get students to be committed to such activities.  That’s the rub.  How to teach students to be better students is the task of most First Year Seminars, but many of my students, even those who had been successful in high school, struggle to meet the intellectual demands or achieve the required self-discipline to be as successful in college.  Engagement is too amorphous a virtue if a student has trouble reading, doesn’t make sense of all the new vocabulary and ways of talking about things, can’t touch-type or navigate Word, struggles with Desire 2 Learn, can’t organize his time, can’t bounce back from failures, or negotiate better hours from his employer. 

Which brings me to the chief issue the research reveals:  causation.  Anyone reading a research report ought to know by now that even strong correlation doesn’t yield a cause, for the simple reason that preceding causes may explain both phenomena, both the dependent and independent variables.  If ACT is right, that high school GPA plus high ACT scores predict college success, we shouldn’t be surprised.  These are students who did well in high school, brought knowledge and skills with them to college, adjusted to the new conditions and moved through successfully.  But what of community college students, many of whom don’t take the ACT, and who often did not do well in high school?

I’ve already mentioned academic engagement as correlating positively with college retention.  It and college GPA earn top awards from Pruett and Absher.  Success (grades) leads to staying in school.  That is helpful information?  How can we use it without giving away A’s and B’s? 

Laskey and Hetzel, while studying students at a four year school but focusing on “at risk” freshmen, point to “agreeableness” and “conscientiousness” as key factors.  The first refers to trusting teachers and doing what they say, the latter seems to be similar to “engagement,” with the same limitations.  Should we only admit agreeable and conscientious students?  Perhaps have a class to teach them these virtues?  How successful do you think we can be?

If students believe they have power over their success and failure, and if they see credible connections between school and their career choice, Luke et al find that their chances of staying in school are high.  But what of those who haven’t figured out what their career goal is? What of those who are unaware of what efficacious behavior in school consists of, or simply are unwilling to commit to behavior that would work?

Cress et al argue that service learning is the key to the kind of engagement that leads to college success.  The programs at the community college level in their review are voluntary.  Do students who participate in community service succeed in school because of their engagement in that service, or were they good students to begin with?  Do students who don’t participate in such endeavors fail to stay in school because they don’t participate, or is something else going on?  Do students who work 20 or more hours a week with a child at home do service learning on the side?  From the 40 page report, it looks like the researchers didn’t ask the question.

Even if we start figuring out what the magic factors are that lead to success—and it seems to me we have more answers than we can use—we still have to figure out where those factors come from.  Stay tuned.

References

Brown, P., Roediger III, H., & McDaniel, M. (2014). Make It Stick: The Science of Successful Learning. Cambridge, MA: Belknap.

Cress, C. M., et al. (2010). A Promising Connection: Increasing College Access and Success through Civic Engagement. Retrieved June 10, 2016, from http://compact.org/resource-posts/a-promising-connection-increasing-college-access-and-success-through-civic-engagement/

Goleman, D. (1985, May 12). Insights into Self-Deception. New York TImes. Retrieved June 10, 2016, from http://www.nytimes.com/1985/05/12/magazine/insights-into-self-deception.html?pagewanted=all.

 Laskey, M. L., & Hetzel, C. J. (2011). Investigating factors related to retention of at-risk college students. The Learning Assistance Review, 16(1), p. 31.

Luke, C., Redekop, F., & Burgin, C. (2015). Psychological Factors in Community College Student Retention. Community College Journal of Research and Practice, 39(3), p. 222-234. doi:10.1080/105589252.2013.803940

Mertes, Scott J. and Richard E. Hoover (2014). Predictors of First-Year Retention in a Community College. Community College Journal of Research  and Practice, 38:7, p. 651-660.  DOI: 10.108/10668926.2012.711143

Pruett, P. S., and Beverly Absher (2015). Factors Influencing Retention of Developmental Education Students in Community Colleges. Delta Kappa Gamma Bulletin, 81(4), p. 32.

Sanders, J. D., & Katz, S. (2013). The Overuse and Misuse of Psychological Testing: Why Less Is More: 1. American Journal of Family Law, 26(4), p. 221.

Westrick, Paul A., et al. (2015). College Performance and Retention: a Meta-Analysis of the Predictive Validities of ACT Scores, High School Grades, and SES.  Educational Assessment, 20:1, p. 23-45, DOI: 10.10880/10527197.2015.997614.

Wyman, F. J. (1997).  A Predictive model of Retention Rate at Regional Two-Year Colleges. Community College Review, 25:1, 29-58.