-  Thirty subjects from a pool of one hundred eighty Washington University undergraduates, ranging from ages 18 to 24, were assigned to six different conditions.  This study was designed to manipulate the retention interval, while examining how the mean proportion of idea units recalled respond to this learning condition.  
-  Both passages that the subjects were required to read were composed of these 30 idea units.  These idea units were used to quantify later retention tests by the proportion of the passage that students would be able to later recall.  The subjects scored one point for each idea unit they recalled correctly out of thirty. 
-  To assure inter-rater reliability, two raters scored the first twenty tests, which yielded a correlation of .95.  The remaining tests were graded by one person. After reading one of the two prose passages, subjects in the SSSS condition underwent repeated study, subjects in the SSST condition underwent a single test, and subjects in the STTT condition underwent repeated tests.  
-  Ninety subjects were given a five minute retention interval before a final recall test, and the remaining ninety subjects followed suit after one week.  To establish a control across these variables, the Experiment 1 passages were used again and counterbalanced across conditions for Experiment 2.  
-  In phase 1 of this experiment the subjects reported an average recall of the passage of approximately seventy percent.  The subjects who studied twice outperformed the subjects who studied once for the five minute retention interval.
-   The one week and two day intervals yielded the opposite results: the tested subjects who studied once outperformed those who studied twice in the final testing phases. In Experiment 1, studying promoted the performance of students for immediate retention tests, while prior testing improved performance on delayed tests 

Experiment 1 Procedure: Test-Enhanced Learning

-  Subjects did not test but studied a passage four times, took one test after studying it three times, or took three tests after studying the material once. They all either took a final test five minutes or one week after their assigned condition. This time the researchers also set out to estimate how many times students in the different conditions actually read each passage.  
-  Keeping the passages constant, the sample size was increased to 180 Washington University undergraduate students.  The researchers were additionally interested in uncovering the students’ own predictions of how well their judgements will correlate with their performance.  
-  After the initial learning session, they included a questionnaire asking subjects to rate how interesting and readable they found the passage and, in the interest of gathering further data, how well they thought they would remember it on a test one week later. 
-  After a week in Experiment 2, sixty one percent of students in the repeated testing condition could recall more than forty percent of students in the repeated-study condition. 
-   Their discovery revealed a pattern between the two experiments: Long-term retention was stronger when the subjects were immediately tested after reading the prose passage as opposed to repeatedly studying the prose passage. The subjects who were re-studying the material experienced poor long-term retention, proving that testing has a powerful effect beyond the process of re-exposing students to material. It was discovered that repeated testing cultivates positive effects for a delayed test, although there are benefits to the repeated study of material shortly after learning

Experiment 2 Procedure: Test-Enhanced Learning

The independent variable: the interval of time between the retention tests and the initial learning condition, and the testing strategy or studying that each student experienced
The dependent variable: the amount of the passage that the student was able to recall

University of Michigan - Ann Arbor

 The testing effect—a phenomenon which suggests that testing enhances later retention—was investigated through a study called “Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention”.  Previous studies analyzing the testing effect fail to do so in an educational context.  Roediger and Karpicke set out to discover this gap in the literature and prove whether or not testing promotes the learning process beyond restudying.  To analyze their hypothesis, the researchers conducted two separate experiments.  Their findings proved that, within educational contexts, testing is a crucial element on both the learning and long-term retention of educational materials than simply the re-studying of these materials.

The Experiment: Test-Enhanced Learning

Roediger, H. L., & Karpicke, J. D. (2006). Test-Enhanced Learning. ​Psychological Science, 17​(3), 249-255. (Page 1)

Learn Before

Related

Learn After