Week 3: Backward Testing Effect

Question 1

Define the backwards testing effect

Answer 1

• The observation that retrieval practice produces superior memory to restudy

Question 2

Outline research looking at student’s study strategies

Answer 2

• Karpicke
• Surveyed college student’s study behaviours - asking them to freely report all study strategies used and ranked in order of frequency
• 84% of students reported passive study such as re-reading notes/textbooks, 55% ranked this as their top strategy!
• Only 11% reported self-testing, and only 1% ranked this as their top strategy!
• Kornell and Bjork
• Gave students the question: ‘if you quiz yourself whilst you study, why do you do so?’
• 68% of these reported it was to figure out how well they had learned the information they were studying
• 18% stated they learn more through self testing compared to rereading.

Question 3

What are other names for the backwards testing effect?

Answer 3

• “Testing effect”
• “Retrieval practice effect”

Question 4

Define ‘retrieval practice’

Answer 4

• The act of attempting to recall information, often involving testing/quizzing

Question 5

What is the simple memory model?

Answer 5

• Encoding → Storage → Retrieval
• Encoding: initial learning phase where participants study information
• Storage: where the information is allocated and stored until retrieval
• Retrieval: where participants are asked to retrieve information from the initial learning phase

Question 6

In an experiment on memory, how can we look at how long the information is stored for?

Answer 6

• Manipulate the retention interval (time between encoding and retrieval)

Question 7

What is the common experimental design associated with the backwards testing effect

Answer 7

• Present participants with study materials to be remembered
• Allocate participants to complete one of two tasks (conditions): restudy: presented the same material and study again OR retrieval practice: tested on memory of what they previously studied.
• Manipulate the retention interval (like: none, 5 min, 2 days, 1 week)
• Administer final test (could be item recognition, free recall, cued recall, MCQ)
• Performance on the final test is the key DV

Question 8

What is a key example of an experiment concerning the backwards-testing effect?

Answer 8

• Roediger and Karpicke
• Participants studied two prose passages on ‘the sun’ and ‘sea otters’
• They studied one passage twice, and the other passage once and then completed an initial free recall test (no feedback)
• Phase 2 (the final test) saw participants completing final free recall for each passage after either 5 min, 2 days or one week later - the retention interval
• Results
• With a short retention interval of 5 mins the restudying condition (studying, studying) had the highest proportion of units recalled
• With longer retention intervals, the study, test condition
• In addition, after phase 1, students thought they would recall more information from the restudy condition 1 week later than the test condition
• Moral of the story: if we want to remember information for longer than 5 min, test-taking is the better strategy!

Question 9

List the 3 key theories underlying the backwards testing effect

Answer 9

• Transfer Appropriate Processing (TAP)
• Retrieval Effort Hypothesis
• Mediator Effectiveness Hypothesis

Question 10

Outline the Transfer Appropriate Processing theory (TAP) of backwards testing effect

Answer 10

• The theory that final test performance should be better when cognitive processes required at encoding and retrieval are the same/similar.
• It predicts that the magnitude to testing effect should be greater when the initial and final tests have the same format (mismatch leads to poorer performance)

Question 11

Outline the Retrieval Effort Hypothesis as a theory of backwards-testing effects

Answer 11

• The Retrieval Effort Hypothesis proposes that retrieval is more difficult than passive study
• So, retrieval practice serves as a ‘desirable difficulty’
• It predicts that the greater the retrieval effort, the greater the benefit of testing
• So, more difficult initial tests (e.g free recall) should produce larger effects than easier tests (MCQ)
• Retrieving after a long delay can produce larger effects than retrieving after short delays

Question 12

Is spaced learning better than massed learning

Answer 12

Yes

Question 13

Define interleaving

Answer 13

• Interleaving is learning categories of things (such as painting styles)
• Interleaving units is where you mix up the order of categories (for example: in the past, I mixed up aggression in psychology with schizophrenia when they were on the same paper, yet different categories)

Question 14

Is blocked units better than interleaving?

Answer 14

• No
• With blocked units it leads to the metacognition that you know something really well when progressing through categories

Question 15

Kornell and Bjork studied interleaving - what did they do?

Answer 15

• They showed paintings with artists names below in a blocked manner (all one artist before progressing) and, in the other condition, showed an interleaved manner (all different artists in progression)
• They were then tested, being required to identify who the artist was when a new painting showed up (by style)
• Interleaved performance was better than blocked performance, when many judged blocked as better than interleaving

Question 16

Why do people have the metacognition that massed learning (cramming) is better than spaced repetition?

Answer 16

• Massed learning creates fluency, misleading them to think that massed learning is better

Question 17

Why do people have the metacognition that blocked learning is better than interleaved learning?

Answer 17

• Block learning creates fluency, so they are mislead to believe this is better

Question 18

What is the flow of the basic paradigm of the backwards testing effect?

Answer 18

• Initial study phase 🢂 Retrieval practice or restudy 🢂 Retention interval 🢂 Final test
• Initial study could include words, word pairs, facts etc
• Retrieval practice could expand to include feedback or none, as well as different test formats
• The final test can be in various formats (another IV in addition to main study condition)

Question 19

How did McDaniel apply the backwards testing effect to real-world scenarios in the classroom?

Answer 19

• His college course had students sit weekly quizzes throughout the course - and then there was an MCQ test at the end of the semester
• The weekly quizzes included a mixture of MC and short-answer questions - and they received feedback
• The control condition was simple exposure: some students were presented with material, yet didn’t sit quizzes on it.
• Results were: those quizzed consistently had a higher proportion correct compared to the control group not quizzed - this effect was most substantial in short-answer.
• This supports the retrieval effort hypothesis (over transfer appropriate processing)

Question 20

What is the Mediator Effectiveness Hypothesis in the backwards testing effect?

Answer 20

• The idea that testing enhances memory performance by activating ‘mediators’ that link cues to targets

Question 21

Aggarwal conducted a meta-analysis on the backwards testing effect (retrieval practice) in the classroom - what was found?

Answer 21

• A wide range of studies were examined, with: different educational levels, content, number of retrieval practice opportunities, timing etc..
• 94% of studies revealed positive effects of retrieval practice
• The majority of effect sizes (57%) were medium or large (if effect size is 0, there is no benefit)

Question 22

Can the backwards testing effect (retrieval practice) improve practical skills?

Answer 22

• Yes
• Medical students completed a course on resuscitation, following simulated cardiac arrest.
• The intervention group completed 3.5 hours of training and 30 minutes of low-stakes testing
• The control group completed 3.5 hours of teaching/training and 30 mins running through scenarios that were covered in the intervention groups test (not practical, just theoretical)
• 2 weeks later, all students completed a final practice test of learning outcomes.
• The final test performance was significantly better in the intervention group than the control group
• So, yes, retrieval practice can be transferrable to practical skills.

Question 23

Can tests protect against stress - outline a study looking at this?

Answer 23

• Stress impairs memory
• Smith et al showed retrieval practice can protect against stress in testing
• Participants had a 2x2 between subjects design - they studied a list of words and images
• They were allocated into either restudy OR free-recall retrieval practice (with no feedback) 3x
• The 2x2 means they were first allocated into one of two IV conditons: stressed vs non-stressed, and further split into encoding group (another IV) which was restudy vs free recall
• After a 24hr retention interval, they either had Stress induction (speech & answering maths questions in front of a crowd) OR non-stressful control tasks (reading, solving problems alone).
• Memory for half of the material was tested 5 minutes into the stress or control task
• Memory for the remaining half was tested 20-minutes later.
• It was found 20 minutes after, the retrieval practice group (whether stressed or non stressed) both scored higher than the study practice groups. (Retrieval practice enhanced recall)
• It was found, for the study group, stress impaired recall.
• Finally, retrieval practiced showed similar recall for stressed and non-stressed participants - so stress did not impact performance much when you do retrieval rather than study.