Does the honor code system work? A randomized study

By Winnie Tam, Centre for University and School Partnership, The Chinese University of Hong Kong

An honor code system comprises rules that govern the conduct of members in an academic community to promote academic integrity, fair play, and discourage cheating. Zhao and colleagues conducted two double-blind randomized studies to investigate the prevalence of cheating under different scenarios during unproctored exams.

Second-year undergraduate students enrolled in the same introductory psychology course at a university in eastern China were recruited for a midterm quiz. Quiz booklets with different forms of honor code reminders were randomly distributed to participants during the quiz. In study 1, students were randomly assigned to one of three conditions: (1) No Reminder condition (n=82); (2) Policy Reminder condition (n=85) with a description of academic honesty policies without mentioning consequences of cheating; and (3) Exemplar Reminder condition (n=93) with two real-life examples illustrating the consequences of academic cheating, including recording the misdemeanor on a student’s file and denial of a degree.

In study 2, participants were randomly assigned to the following conditions: (a) Consequence Reminder only (n=128); (b) Trust Exam and Consequence reminder (n=125); (c) Policy and Consequence reminder (n=129); and (d) Exemplar Reminder (n=128). In study 2, all conditions included a description of the negative consequence of academic cheating. Moreover, conditions (b) to (d) informed students that the quiz was a trust exam. Researchers measured the extent of cheating by counting the number of target questions answered correctly. The cheating rate indicated whether a student had cheated or not. The findings of both studies are shown below.

In study 1, the cheating rates were: No Reminder condition 54.2%; Policy Reminder condition 31.8%; and Exemplar Reminder condition 22.6%. In study 2, the cheating rates were: Consequence Reminder 26.6%; Trust Exam/Consequence Reminder 24.0%; Policy and Consequence Reminder 17.1%; and Exemplar Reminder 15.6%.

Unproctored exams without an honor code reminder or consequence of cheating reminder had the highest cheating rate (54.2%). Reminding students of the trust exam policy significantly reduced cheating extent compared to the No Reminder condition. However, providing negative consequences for academic dishonesty, through adding a statement or telling actual examples, had an even greater effect in reducing cheating. The authors noted that mentioning negative consequences of academic cheating may imply a threat, contradicting the principles of the honor code system. However, the research results provided evidence that this strategy effectively reduces cheating during unproctored exams. Despite various reminders, approximately 20% of students still cheated on average. The authors speculated that the honor code system has not yet become fully integrated into university academic culture.

How to conduct a more effective flipped class for English language learning

By Winnie Tam, Centre for University and School Partnership, The Chinese University of Hong Kong

The flipped classroom approach, where students learn new content via video lectures outside class and use class time for interactive applications, is gaining popularity in higher education.  A meta-analysis by Ni and colleagues, published in the International Journal of Educational Research, explored the impact of the flipped classroom approach on college students’ English learning outcomes.

The meta-analysis included 24 studies conducted from 2000 to 2020, yielding 31 effect sizes. These studies focused on the use of the flipped classroom approach in English language learning, wherein instructional videos and other technological tools were employed prior to class sessions to enhance college students’ performance. All studies followed a pre-test, post-test control group design, comparing students who experienced the flipped classroom with those taught using traditional methods. The analysis revealed that flipped classrooms had a positive  overall effect on learning outcomes (ES = +0.68). Interventions that included a review of pre-class preparation during class had significantly higher effects (ES = +0.92) than interventions without such reviews (ES = +0.38). A significant difference was observed among three types of constructive learning models: self-learning (k=3, ES = +0.22), cooperative learning (ES = +0.58), and a blend of the two (ES = +1.10). The findings of this meta-analysis offer valuable insights into the conditions under which flipped classrooms are more effective in improving students’ English language learning outcomes.

The long-term impact of KIPP on college enrollment and persistence

By Nathan Storey, Johns Hopkins University

The Knowledge is Power Program (KIPP) network of public elementary, middle, and high school-level charter schools is the nation’s largest. While there is data on student academic achievement in KIPP schools, there is less research on its impact on longer-term outcomes, such as enrollment, persistence, and attainment at the college level. A recent study by Demers and colleagues seeks to change that.

Using a randomized controlled trial design ensuring that students offered admission to a KIPP middle school were similar to those not receiving an offer on characteristics including prior test scores, motivation levels, and parental support, the researchers sought to examine the impact of KIPP middle schools on student enrollment and persistence in a four-year college during the first three years after high school graduation. Demers and her colleagues tracked 2,066 students who partook in a lottery  to enter grade 5 or grade 6 at 21 oversubscribed KIPP middle schools across multiple states during  the 2008–2009 to  2011–2012 school years..

The study found that KIPP middle schools had a positive but not statistically significant effect on college enrollment (p = +0.12), when comparing those who received an admission offer to those who did not. In addition, students’ college persistence rate (p = +0.78) and graduation rate (p =+ 0.99) were similar to those who did not receive KIPP middle school admissions offers. However, the authors noted that students who attended both a KIPP middle and high school were 30.5 percentage points more likely to enroll in a four-year college (p < 0.000), almost 20 percentage points more likely to persist in college for at least three years (p < 0.000), and 18.9 percentage points more likely to graduate from a four-year college (p = 0.025). They noted that these college attainment rates exceed the gaps among Black and Hispanic students compared to White students in the United States.

These findings suggest that it is not enough to just receive an admission offer to KIPP middle school, or to just attend a KIPP middle school. The stronger effects of attending a KIPP middle and high school are potentially connected to college preparatory culture development efforts and college-related supports that are focused more centrally at KIPP high schools. These include rigorous coursework for all students, college counselors paired with every KIPP high school student, and financial aid application and precollege summer program identification supports.

The benefit of learning by teaching

By Winnie Tam, Centre for University and School Partnership, The Chinese University of Hong Kong

Learning-by-teaching is a pedagogical approach where  learners teach the material  they’ve studied to others.   Wang and colleagues conducted an experiment with 96 college students from a university in central China comparing  three versions of learning-by-teaching. Participants studied a 2-minute video on chemical synaptic transmission for 9 minutes and prepared a brief lesson of less than 5 minutes while randomly assigned to one of three conditions: (1) teach-to-camera – teach to an imaginary audience by creating a video lecture; (2) teach-to-student – teach to an audience face-to-face; (3) teach-to-group – teach to seven people physically present in the room. Audiences in the latter two conditions provided no feedback. Data collecting involved learning outcomes scores, self-reported questionnaires, pulse rate, and teaching process recordings.  Results indicated that:

• Students in the teach-to-camera condition performed better than those in the other two conditions in terms of generative processing(more idea units, elaborations, and monitoring statements in their explanations).
• The teach-to-camera participants outperformed the teach-to-group students  in the retention test and the transfer test, and scored better than those in the teach-to-student condition on the retention test.
• The teach-to-camera participants also reported significantly lower social presence and pulse rate than the two other groups, and perceived lower state anxiety, teaching difficulty, and cognitive load in teaching than those in the teach-to-group condition.

The authors suggested that the better performance observed in the teach-to-camera arm of learning-by-teaching might stem from the absence of a live audience. This absence reduced social presence, thereby minimizing extraneous processing (such as anxiety, cognitive load, and increased pulse rate) and enhancing generative processing.