Absences under a revised policy

Summary: The mean number of absences was found to be statistically unchanged under the revised attendance policy for two courses. A third course showed a statistically significant increase in the mean number of absences.

Note and policy background

The following is a note on the mean number of absences in three courses at the College of Micronesia-FSM across multiple terms. The three courses are MS 150 Statistics (six terms), SC 130 Physical Science (five terms), and SC/SS 115 Ethnobotany (four terms). The background and genesis for this note is the shift from a college-wide catalog mandated attendance policy to individual attendance policies set by faculty.

The policy in effect up until spring 2009 was:

Students who are absent for more than six MWF classes, or more than four TTh or summer classes, will automatically be dropped from the course. The total missed classes include unexcused and excused absences, such as sickness, funerals, and any other circumstances.

The policy in effect fall 2009 reads:

Regular and prompt class attendance is expected of all students. It shall be the student's responsibility to inform the instructor(s) of anticipated or unavoidable absences and to make up work missed as a result of absences. Mandatory attendance is at the discretion of the instructor provided the conditions for the attendance are included in the course syllabus and communicated to the students on the first day of class.

I chose to retain the spirit of the college policy, using that policy minus the words "and excused". I allowed for excused absences that would not be counted against the absence limit. Excused absences would include documented medical absence or absence due to documented academically related travel such as to a conference.

I also included, as I have always done, that a late is one-third of an absence. Thus three late arrivals is equal to one absence. In some instances, when the student has arrived half-way through class, I have recorded a half-absence. The result is the possibility of fractional absences in my courses.

Chart explanation

For each term I calculated the mean number of absences and the 95% confidence interval for that mean number of absences. The chart further below uses course and term encoding abbrevations.

The prefix s refers to statistics. Statistics is a MWF course.
The prefix ps refers to physical science. The physical science data includes only MWF attendance in the lecture section for purposes of comparison to statistics.
The prefix e refers to ethnobotany. Ethnobotany is a Tuesday, Thursday non-lab science course.

The suffix 63 is the fall 2006 term.
The suffix 71 is the spring 2007 term.
The suffix 73 is the fall 2007 term.
The suffix 81 is the spring 2008 term.
The suffix 83 is the fall 2008 term.
The suffix 91 is the spring 2009 term.
The suffix 93 is the fall 2009 term.

Thus ps93 is the physical science fall 2009 term data.

The courses are arranged in the order statistics, physical science, and ethnobotany from left to right. The y-axis is the number of absences. The horizontal ticks on the vertical bars are the mean number of absences for that course in that particular term. The means in the chart include both excused and unexcused absences for fall 2009, the only term in which a distinction was made. The vertical bars are the extent of the 95% confidence interval for the mean. Where vertical bars overlap ticks between two terms, those two means are not statistically significantly separated.

Click on the chart for an enlarged view.

Discussion

MS 150 Statistics not only showed no statistically significant change in the number of absences. Under the new policy, the mean of 5.99 absences is statistically identical to the long term average of 6.04 absences in the course. The new mean is the third lowest in the history of the course. This is particularly noteworthy as the course included an extra section this term, increasing course enrollment by 60% over the prior term.

SC 130 Physical Science showed a small mathematical increase with a mean of 5.44 absences fall 2009. The four term average was 4.7 absences. The mean number of absences for fall 2009, however, was not statistically separated from prior term mean absenteeism. Were the increase to mark the start of a trend, then the separation could be significant within a term or two. There is no way to determine whether a trend exists based on a single term. For now, the new policy cannot be said to have had impact.

SC/SS 115 Ethnobotany saw a statistically significant increase in the mean number of absences. Mean absenteeism more than doubled from a three prior term average of 2.92 to 6.18 absences. While there are some identifiable underlying factors, such as 15 of 30 students being absent on day one of the course, in general there is no single cause for the high mean number of absences. Sickness, injuries, even hospitalization seemed to strike more students than has occurred in prior terms. Some students had a combination of the full limit of unexcused absences plus other excused absences. In this one course, students who had been ill still "racked up" the full complement of allowed unexcused absences. The time slot, 3:30 on a Tuesday or Thursday, might also be a factor.

Conclusions

The new attendance policy, at least as deployed in my courses, has had no appreciable impact in two of the courses. The increase in the mean number of absences in the third course, ethnobotany, may or may not be directly connected to the new policy. Absenteeism, however, was fairly constant in that course prior to this term. While many other variables are involved - a whole new class of students, the passage of time - the new policy as implemented may have contributed to the rise in absences in that one course.

As deployed in the manner I did, the new attendance policy has not had a broad negative impact on attendance that some members of faculty had been concerned might occur.