Testing Effects in Long Term Memory | 1

Testing Effects in Long Term Memory

Lissett Liliana Martinez Mendoza

Psych 240: Cognitive Psychology

Dr. Nick Ellis

Natasha Vernooij

December 7, 2020
 Testing Effects in Long Term Memory | 2

Abstract

Testing is a way to know what students learn, but is underutilized as a learning tool. Multiple

studies have shown that testing is one of the most effective methods for learning and has far

more benefits than just to evaluate. This study, Test-Enhanced Learning by Roediger and

Karpicke at Washington University in St. Louis, was done in an educational context with

undergraduate students using a free-recall test of prose passages. They hypothesized that test

taking had long term learning benefits that were not obtained by studying. Additionally, the

experiments showed increasing testing events increases long term recall. The study had two

experiments. The first experiment demonstrated that restudying gives better results in the short

term, but testing provides better long term results by specifically promoting deep learning.

Roediger and Karpicke’s work also discusses implications of the testing effect in cognitive

psychology.
 Testing Effects in Long Term Memory | 3

Testing Effects in Long Term Memory

The methodology of Roediger and Karpicke was to create two very specific experiments

that now provide a solid base for renewed examination of the theoretical and practical

implications of efficient study methods. Conversely, in addition to solid evidence for effective

study methods, they also make us consider the costs of poor study habits and to ask questions

about why these habits continue while exploring topics of further study. There were also other

conclusions about studying vs. testing such as forgetting, insecurity with the material, interest in

the material, and overconfidence.

Experiment 1

Experiment 1 studied testing effects in 120 subjects of ages 18 to 24. The subjects read

two prose passages of 256- 275 words long. The experiment was a 2 x 3 mixed-factorial design,

meaning that the independent variables have different levels. The independent variables were

the learning conditions (restudy and practice test) and delays on the final test (5 minutes, 2 days

and 1week). The score (the number of recall units) was the dependent variable. The procedure

was in two phases. In phase one, the subjects in small groups were tested in two sessions, each

time divided into four periods of 7 minutes. Each group would either have to study one of the

passages for 7 minutes, restudy or take a recall test. During the test, the subjects wrote down

what they remembered from the passage. In between periods, the subjects solved multiplication

problems for two minutes. In the second phase of the experiment, subjects were given a final

test either in 5 minutes, 2 days or 1 week. The test was the same as in phase one, but the

duration was 10 minutes. The subjects were told to recall the passages from phase one and write

them down.
 Testing Effects in Long Term Memory | 4

Results

The results of experiment 1 were that the average recall in the initial 7 minute test was

20.9 idea units equal to 70% of the passage. In the final test, the 5 minute delay showed that

subjects from the pure-study group had better recall at the short delay, however, the recall

decreased as the test interval increased. The opposite happened in the final test of long-term

retention: subjects who took a test instead of restudying showed better long term retention.

Experiment 2

In the second experiment, Roediger and Karpicke’s goal was to prove that testing

enhances learning in delayed tests. This experiment was similar to experiment 1. The subjects

only read one passage, then, either only restudied, studied the passage multiple times and took a

test, or studied the passage once and took multiple tests. In addition to this, the subjects received

a questionnaire to rate their experience, and a final test was given. This experiment was a 3 x 2

between-subjects factorial design and the independent variables were the learning conditions

(three levels: SSSS, SSST, or STTT), and the testing delay (retention interval 5 minutes or 1

week), and the dependent variable was the score (the number of recall units). In phase 1, the

subjects studied one passage in four periods: the learning conditions SSSS (only study), SSST

(studied three times and took a test) and STTT (studied once and took three tests). In the second

phase, there were the same learning conditions, but subjects took a final test either 5 mins or 1

week later.
 Testing Effects in Long Term Memory | 5

Results

The results in experiment 2 showed that the subjects had better long-term retention, as

testing increased, than the restudy condition. The score in the initial test in condition STTT was

20.9, 21.2, and 21.1 equal to 70% of the passage whereas in condition SSST the recall was 23.1

(77%) (at the 5 minute interval). The study also showed that as studying goes up interest goes

down and when testing increases interest increases. In addition, however, it showed with

repeated studying confidence increases but with repeated testing confidence decreases. In the

final test, there was a correlation with experiment 1 in the 5 minute recall the condition SSSS

showed 83%, SSST 78%, and STTT only 71% and the 1 week results showed the opposite trend,

SSSS 40%, SSST 56% and the STTT 61% . The results of the delayed 1 week test reflected the

benefits in testing, and it also showed a main effect on retention interval indicating that there is

more forgetting in the only study condition compared with the rest of the conditions.

Context of the research

The results in this study showed the same pattern as previous studies on testing, for

example, in 2010, Butler’s experiment demonstrated that taking tests enhances learning. Clearly

it provides proof about a more effective retrieval method to improve long term recall (Ellis,

2020). There is a history of interest in learning efficiently and there are known good and bad

habits. For example, there are good habits such as the desirable difficulties, testing, generating,

spacing, interleaving, pre-reading, note taking and flash cards, among others (Ellis, 2020).

There are also bad habits, like restudying, going over notes, mass studying and studying in the

same place. This study was needed to show there are more efficient strategies and showed

students who took a test did better on longer delay test intervals because every time they tried to
 Testing Effects in Long Term Memory | 6

retrieve the information they made a conscious effort to remember the information (Ellis, 2020).

This meant that the repeated study group engaged in shallow processing whereas the testing

group engaged in deep processing. Understanding the research on testing effects makes a very

strong argument for decreasing your comfort level when studying and increasing your discomfort

level. Or, stopping unproductive habits that give a positive feeling of overconfidence and

replacing it with the better habits of studying that may give the negative feeling of being

insecure. One of the reasons why the repeated study group felt secure is because there is a point

where material becomes familiar due to repeated exposure, and this exposure gave the false

sense of learning when in reality they were only grasping the material at a surface level (Ellis,

2020).

Theoretical implications

One of the theories that these researchers refer to is transfer appropriate processing,

which is the idea that recall is better when the environmental conditions are the same at retrieval

as when the information was encoded. This was one of the explanations given by the researchers

for the subjects who achieved greater recall. Apparently their recall was better because the

subjects restudying had the same cues when they took the test. This is also supported by

encoding specificity which supports the idea that memory would be better if the same cues are

present in encoding and retrieval (Ellis, 2020). Another implication is the practice of the skills

when learning. This means that each time the subjects are taking tests they are practicing the

skills they need for retrieval. Roediger and Karpicke mentioned that there is probably a

relationship between the memory paths and cues with the benefits of testing which could show

that testing helps to strengthen the neural pathways that were created when the material was

presented, so each test strengthens these pathways.

Practical implications

The proof of the testing effect (Roediger & Karpicke, 2006) could help the academic

system since students and teachers have been trying to find the best study methods to apply in

educational settings. The knowledge that testing enhances memory can help schools to focus on

methods that can help students achieve more than just academic success by promoting deep

learning. More importantly, there should be a use of the effective methods recommended by

cognitive scientists and supported by research such as spacing, generating, group study,

flashcards, self-testing, Cornell notes, previewing things in order to prime yourself and reading

with purpose and thinking about what you want out of a reading before reading to begin to code

into existing knowledge (Ellis, 2020). Another practical implication is to create or practice more

one on one and group developed testing, and work on hints that have deep learning elements and

an organizing principle attached to them.

Conclusion

This research supported the hypothesis that testing promotes deep learning. It also made

it difficult to overlook the problems of shallow learning, which can have consequences for

society. For example, it made it possible to think some people are getting an education without

learning how to learn, or worse, have limited deep learning experiences in school. This study

showed that the stronger proven method of testing and strategies from cognitive psychology

should influence schools and students to alter their habits. Even further, we can use this research

to consider how our reward systems, our response to stress, and other factors influence our

choices to continue unproductive habits.

References

Ellis, N. (2020, September 3). The Cognitive Psychology of Optimal Study in College [Lecture

recording]. [email protected]. https://wolverineaccess.umich.edu/

Roediger, H. L., & Karpicke, J. D. (2006). Test-Enhanced Learning: Taking Memory Tests

Improves Long-Term Retention. Psychological Science, 17(3), 249–255.

https://doi.org/10.1111/j.1467-9280.2006.01693.x

What Is Transfer Appropiate Processing? (2018, February 22). Retrieved November 3, 2020,

from https:www.mysimpleshow.com/transfer-appropiate-processing/