Virtual School Meanderings

March 22, 2016

SITE 2016 – Does Class Size Matter in Online K-12 classes?

As I mentioned in the entry entitled SITE 2016 And K-12 Online Learning, the the Society for Information Technology and Teacher Education (SITE) 2016 annual conference is occurring in Savannah, Georgia this week, and SITE is home to the K-12 Online Learning SIG.  That means that I will be blogging many of the sessions throughout the week.  The fifth session I am blogging is:

Does Class Size Matter in Online K-12 classes?

  1. Chin-Hsi Lin, Michigan State University, United States
  2. Binbin Zheng, Michigan State University, United States
  3. Joe Freidhoff, Michigan Virtual Learning Research Institute, United States

Tuesday, March 22 11:30-11:50 AM in Sloane View on map

Discuss  Download Paper

The effect of class size in traditional mortar-and-brick classroom has been believed to be a factor affecting students’ achievement, but the effect of class size in online education has not yet received much attention. Using a dataset of 20,540 enrollment records from a state virtual school in 2012-2013, this study examined the relationship between class size and students’ learning outcomes. The results of Hierarchical Linear Modeling (HLM) suggest a linear relationship, indicating economy of scale for online courses.

ID
48251
Type
Brief Paper
Topics
Distance/Flexible Education K-12 Online Learning

Chin Hsi began with some of the differences or benefits of smaller class sizes, and the research into effects of class size in the traditional classroom.  Chin Hsi transitioned to some of the research into what was the ideal class size – both in face-to-face and online environments.  However, much of this research is based on perceptions, as opposed to outcomes.

Chin Hsi’s data was from a supplemental virtual school with 21,253 enrollments from 12,445 students in 233 different courses in six subjects.  He used HLM, which required him to exclude students who had multiple enrollments (or at least he excluded all but their first record) – as HLM requires a level of independence between instances.

  • Level 1 – students enrollment record
  • Level 2 – each class section
  • Model A: unconditional model – explains 24.9% of the variance
  • Model B: add class size (i.e., linear) – statistically significant
  • Model C: add class size quadratic term (i.e., non-linear) – statistically significant (after 45 students it did start to decrease)
  • Model D: add class size cubic term (i.e., non-linear, non quadratic) – not statistically significant (but based on the data from model C it does suggest the relationship may be non-linear)
  • Model E: fractional polynomial analysis – statistically significant (smallest classes had lower performance, as size increased so did performance, until the class size reached 42-45 students at which point it began to decrease)

When looking at specific subjects, Chin Hsi found:

  • English language arts – class size was found to be statistically insignificant
  • foreign languages – class size was found to be statistically insignificant
  • science – class size was found to be statistically insignificant / but there was an increase when class size increased from 1-5 to larger
  • math – statistically significant / with a significant decrease after about 40 students
  • social science – statistically significant / with a significant decrease after about 43-44 students
  • other subjects – class size was found to be statistically insignificant / but there was an increase when class size increased from 1-5 to larger

While Chin Hsi didn’t focus on this, as best I could tell this was based on students that had completed the course (i.e., were retained).  I don’t believe it reflects those that did not receive a grade (i.e., dropped out), which class size (and all of the things that entails) could play a significant impact.

9 Comments »

  1. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. - The ABM Media — October 21, 2020 @ 10:36 am | Reply

  2. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. | EdSurge News — October 21, 2020 @ 10:39 am | Reply

  3. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. - EdSurge News - Country Highlights — October 21, 2020 @ 10:59 am | Reply

  4. […] college students, for instance, smaller class sizes of round 15 college students are preferable. In a separate study, Lin used fashions to find out that college students’ remaining grades started to undergo as soon […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. – worldnewsas.com — October 21, 2020 @ 11:21 am | Reply

  5. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. – Family Educational Learning Pod's — October 21, 2020 @ 1:34 pm | Reply

  6. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. - EdSurge News - NWS100 — October 21, 2020 @ 4:50 pm | Reply

  7. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. | EduTalks — October 22, 2020 @ 4:01 am | Reply

  8. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. – Book Of Celebs — October 22, 2020 @ 2:28 pm | Reply

  9. […] between students, for example, smaller class sizes of around 15 students are preferable. In a separate study, Lin used models to determine that students’ final grades began to suffer once class sizes […]

    Pingback by K-12 Class Sizes Have Ballooned With Online Learning. It’s Not a Good Thing. – English News — November 12, 2020 @ 10:17 am | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Blog at WordPress.com.

%d bloggers like this: