zyBooks improve student learning outcomes


Alex Edgcomb

zyBooks software and content developer

Post-doctoral researcher, Univ of California, Riverside

I completed my PhD in CS earlier this year in the area of assistive monitoring of the elderly in at-home environments. My dissertation was on a camera-based algorithm for detecting when an elderly person falls down. Seemingly light years from education research.

However, on the PhD trek, I learned that I love teaching, both interactions with students and developing learning materials. I also learned that STEM retention was (and still is) a significant and unsolved problem. So, in my last 2 years of the PhD, I started investigating how to improve STEM retention.

A key aspect of STEM retention is improving student learning outcomes.

Recently, I conducted two studies on improving student learning outcomes: one focused on a single lesson; the other had a course-wide lense. Both studies found zyBooks significantly improve student learning outcomes over traditional textbooks/eTextbooks.

First, I’ll share the two studies, then I’ll share my research next steps.

With a single lesson, zyBook’s improved least-prepared students by 64%

I measured the effectiveness of a zyBook lesson to a traditional eTextbook lesson. I even added a programming environment along with the traditional eTextbook lesson so students could practice the lesson, if the student wished. 136 students in an intro. to programming class participated. Each student completed 4 steps:

  1. a pre-lesson quiz
  2. a lesson (randomly assigned either zyBook or eTextbook)
  3. a post-lesson quiz
  4. a follow-up survey

RESULT: The lowest-quartile of students on the pre-lesson quiz (a.k.a. least-prepared students) improved 64% more with zyBooks than the traditional eTextbook. Improvement score is post-lesson quiz minus pre-lesson quiz.

Static means traditional textbook. Interactive means zyBook.

-5 Why did students improve so much more with a zyBook? Simple: Higher level of engagement with visualizations and immediate feedback. Students self-reported 10.4% higher level of engagement with the zyBook. Also, students voluntarily spent 2x as long with the zyBook, even though, the zyBook had 1/2 as many words. One student comment about the zyBook was: “I can learn by practicing and auto-check the answer right away … the activities are fun and helpful.”

Across 4 courses at 3 Universities, zyBooks improved student grades by 1/4 letter

I retrospectively compared the same course across two offerings. The first offering was taught using a traditional textbook; the second offering using a zyBook. Both offerings were taught:

  1. by the same instructor
  2. during the same semester of the year, e.g., Spring 2013 and Spring 2014
  3. within 2 years of each other

I conducted the same comparison across 4 courses at 3 Universities: Univ. of Arizona, Univ. of Michigan, and Univ. of California at Davis. A total of 1,945 students were part of the comparisons. RESULT: zyBook students had more As and Bs, fewer Ds and Fs. The Univ. of Ariz. CS2 students improved pass rate by 8.6%. The UC Davis ENG1 saw students improve letter grades by 3/4 a letter.

Static means traditional textbook. Interactive means zyBook.

 -2 -1  -3     -4

Then, I combined the comparisons.

RESULT: Across the board, zyBooks very-significantly improved student learning outcomes. Students using the zyBook earned a 1/4 higher letter grade than students using the traditional textbook. Exams and projects significantly improved with zyBooks.

Exam Projects Class score Class letter grade
zyBook improvement % (higher is better) 13.6% 7.4% 14.3% 12.6%
Significance value (p < 0.05 is significant) p < 0.0001 p < 0.0001 p < 0.0001 p < 0.0001

Research next steps: Why do zyBooks work so well, and how can we optimize?

The two studies indicate that zyBooks improve student outcomes. But why?

zyBooks contain many content types: animations, various question formats (including true/false, multiple choice, and short answer), and interactive tools. I want to know which of these content types improve student outcomes most. Between the content types, there’s a trade-off between effort to develop and impact on student outcome, but at this point, only the effort to develop is known.

I also want to know what the optimal usage of each content type is. For example, when is a multiple choice appropriate vs. when is a short answer question appropriate. Many eTextbooks predominantly (if not exclusively) use multiple choice questions, but does that benefit the student most in all cases? Intuitively, no. Multiple choice are inherently easier because they present all options. Seems that a student would learn more by being incrementally challenged to the point of answering short answer questions.

Also, how can student outcomes be further improved? One direction could be practice quizzes, including quizzes before the chapter has been started (see NY Times on how flunking exams helps).

A later step is exploring other methods for improving STEM retention. One possibility is showing the relevance of each topic to daily life. There are many questions to answer, but the road looks promising. Similar questions and more are being investigated globally, and investigated on different fronts. What a thrill to be a part of that.

Visit us at zybooks.com

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s