learning to learn key studies Flashcards
Bjork (1994)
(desirably) difficult learning = enduring learning
Yunker and Yunker (2003)
explored ratings/final marks relationship between the two courses with the same students but different instructors
- after controlling for student ability, higher ratings in introductory accounting were negatively correlated with grades in intermediate accounting
- found higher ratings for instructors performed worse
Karpicke et al (2009)
surveyed college students’ study behaviours (N=177). Students freely reported all study strategies that they had used, and ranked them in order of frequency used
- re-reading used by 84%, 55% ranked it as top strategy
- self-testing used by 11%, 1% ranked it as top strategy
Cornell and Bjork (2007)
“if you quick yourself while studying, why?”
-68% - to figure out how well information has been learnt
- 18% - learn more through testing than re-reading
- 4% - quizzing is more enjoyable
- 9% - don’t usually quiz self
Reodiger and Karpicke (2006)
landmark paper.
experiment 1, phase 1 - participants studies 2 prose passages on “the sun” and “sea otters”. they studies one passage twice, and the other passage once then completed and initial free recall test
- phase 2 - participants completed final free recall tests for each passage either 5 min, 3 days or 1 week later.
- results - very short retention intervals, restudying > test. longer retention intervals, test> restudying.
- if we want to remember longer, test taking is better
McDaniel et al (2007)
real college course, students took weekly quizzes or given read only information. quizzes were either multiple choice questions or short answer
- quizzing, but not restudying enhanced final test performance relative to control material
- short answer quizzes were more beneficial than multiple choice questions
Agarwal et al (2021)
meta-analysis of 50 experiments examining effects of retrieval practice in classroom
-94% of studies revealed positive effects of retrieval practice
-majority of effect sized (57%) were medium or large
Kromann et al (2009)
medical students completed a resuscitation course following simulated cardiac arrest
-intervention group completed 3.5 hrs teaching /training. then 30 mins low stakes testing
- control completed 3.5 hrs teaching/training, ten 30 mins of scenarios
- 2 weeks later, final practice of learning outcomes
- final test performance was significantly better in intervention group than control group
Smith et al (2016)
participants studies list of words and images.
-restudy or free-recall
- 24 hours later -stress induction or non-stressful control tasks
- overall retrieval practice enhanced recall
- study group: stress impaired recall
-retrieval practice group: similar recall for stressed and non-stressed participants
Szpunar et al (2008)
taking a test after lists 1-4, improved learning and recall of list 5, relative to a control task or restudying lists 1-4 - the forwards testing effect.
- test taking led to fewer intrusions from the previous lists than the other conditions
- benefit of testing on cumulative test probably partially reflects the backwards testing effect
Szpunar et al (2013)
students watched an online lecture in 4 segments
- participants completed either test, restudy phase or control task after first 3 segments
- all tested after the 4th segment
- participants asked whether they mind-wandered during segments
-experimenter looked at students’ notes from video
- interim testing improved learning of 4 segment, reduced mind-wandering and increased note-taking
Jing et al (2016)
undergraduate students watched 40 min lecture with interim testing or restudying
-participants asked what they were thinking about during the lecture
-mind-wandering did to significantly differ between the two groups
- although the interim testing group reported mind-wandering more closely related to the lecture than the restudy group
Yang et al (2019)
participants studied either face-name or Swahili-English pairs in lists 1-3. all participants studies face-name pairs in list 4
- interim testing improved final list 4 recall even when lists 1-3 and list 4 material type differed.
- forward testing effect is transferable to different stimuli
Kornell et al (2009)
participants studied weakly related word pairs. no exposure to answers before the pretest
- participants guesses mostly incorrect, any correct guesses removed from data set
- testing improved learning, even when all answers were wrong
- errors fostered learning
Grimaldi and Karpicke (2012)
same as Kornell et al (2009), however they used weakly related and unrelated pairs.
- pretesting improved recall of related pairs, not unrelated pairs
Seabrooke et al (2019)
Participants gave higher motivation ratings to learn facts they had guessed than not guessed
Butterfield and Metcalfe (2001)
participants answered questions and rated their confidence for each. feedback was provided after each answer
-5 min retention interval
- cued recall final test
- high confidence errors more likely to be corrected than low-confidence errors
Butterfield and Metcalfe (2006)
participants completed a hypercorrection task. also asked to detect soft tones.
-participants missed more tones that were presented with feedback following high-confidence errors than low-confidence errors
-suggests participants’ attention was captured by the feedback
Ebbinghaus (1885)
memorised nonsense syllables to the ticking of a metronome and then attempted to recall them later.
- massed learning day 1: 68 massed repetitions, day 2: 7 additional repetitions were needed for 1 perfect recital
-spaced learning: day 1: 38 spaced repetitions, day 2: 7 additional repetitions were needed for 1 perfect recital
-with any considerable number of repetitions a suitable distribution of them over a space of time is decidedly more advantageous than massing of them a single time
Ruch (1928)
reviewed dozens of studies on the spacing effect and concluded that the spacing effect was robust
Greeno (1964)
participants studies word-digit pairs, then shown words and asked to name associated digit. 3 presentations of each pair with feedback. presentations spaced (15 intervening items) or massed (0-1 intervening item)
-spaced shown to be more effective
Bird (2010)
found that longer spacing gaps improved English-learning adults’ understanding of subtle grammatical rules
Rohrer and Taylor (2006)
students required to find the number of permutations of a sequence of items with at least one repeated item. either practiced all at once (massed) or the problems were split over two sessions (spaced)
Baddeley and Longman (1978)
postmen trained over several sessions to type aloha-numeric code material using a convertible typewrite
-learning: one 1hr session per day over 12 weeks > two 2 hr sessions per day over 3 weeks
-metacognition: opposite results were obtained for questions : “how satisfactory did you find your training?”
Rohrer and Taylor (2007)
students practiced computing the geometrical volume of four differently shaped objects
-blocking (massed) leads to good short-term (practice performance, thought they mastered material, but it is bas in the long-term
Cepeda et al (2008)
tested 26 combinations of spacing and retention intervals in massive online study
-participants learned obscure facts, has a review session, and wrote a recognition and recall test
Higham et al (2023)
the number of times of practice sessions increases proportions of correct answered and the retrieving practice type also increases correct answers
Rawson, Vaughn, Walsh and Dunlosky (2018)
48 Lithuanian-english word pairs. one initial study per item
- relearning with dropout and mastery
-criterion level: 1 or 3
-corrective feedback following errors
- 4 further relearning sessions each separated by one week
- 3 weeks after 4th relearning = 77%
Tucking an Thomson (1973)
participants failed to recognise several words (in step 3) that they were able to recall in (step 4): light; blue, baby.
- participants were trying to recognise targets generated in the context of strongly related cue words. then weakly related cue words were presented again.
-recall was successful but recognition failed
Morris, Bransford and Franks (1977)
participants encoded nouns with an orienting task that either; encouraged deep (semantic) processing or encouraged shallow (phonemic) processing
-deep semantic encoding leads to better standard recognition but leads to worse rhyme recognition
-if the test is sensitive tot he type of processing that occurs during encoding, then memory will be good
Jacoby (1983)
also demonstrated that memory is dependent not just on how the information is encoded (e.g. Deep vs shallow), but now whether testing draws on encoding processes
- generation effect
- like levels of processing, only desires the encoding; ignores how memory is tested
-generating vs reading words leads to better standard recognition but worse perceptual identification
Smith et al (1978)
investigated people’s memory for word lists that were either studied twice in the same environment context, or into different contexts, they were then tested in.a third neutral context - perry schools context mason hall context, and neutral context. the same context was better than different contexts for recall
Imundo et al (2021)
looked at the effect of testing (vs restudying) in the same vs different contexts.
-encoding variability hurts performance if practice testing
- same context: 36% recall > varied context: 17% recall
Fisher and Geiselman (1992)
cognitive interview, used in forensics to maximise retrieval with eye witness and victims of crime. 4 stages
1. mentally reinstate context
2. recall events in reverse order
3. report everything
4. describe events from someone else’s point of view
- the purpose was to increase the overlap between the initial learning context and current retrieval context
Mcdaniel et al (2007)
web based university course on brain and behaviour. assigned weekly reading and then practice with; mc quizzing, short answer quizzing, rereading. corrective feedback was provided
-suggests sort-answer is best. some other studies disagree
Marin et al (2009)
had participants answering SAT 11 MC test questions on biology, chemistry, world history, US history
-formula scored, option to omit response
-filler task
- answer short answer questions - 40 items from earlier test or 40 new items
-undergraduates performed better of SA. high school juniors fell for MC lure
Little et al (2019)
participants first completed an online MC practice test with general knowledge questions
-elimination testing to encourage processing of all lures
after distractor task, completed cued-recall test
-previously tested items, related questions new questions
-repeated did best, however whether practice with repeated items facilitated later test performance depends on whether the final test is in cued recall of MC format
Alamri and Higham (2022)
group 1: practice test = MC; final test = cued-recall.
group 2: practice test = MC; final test MC
- replicated little et al’s results. most errors due to participants selecting corrective feedback from practice test. participants believed related questions were repeated questions - false recognition
Kelley et al (2019)
peer wise involves two effective learning techniques: generation - generating information leads to better memory than reading. retrieval practice - retrieving information leads to better long term retention than restudying
- 40 students enrolled in cognitive psychology course
- generation technique performed best
Sparrow et al (2011)
participants presented with easy trivia questions in one block, and hard trivia questions in another block. after each block participants completed a modified stoop task - hard questions had longer reaction time
Sparrow et al (2011) - recognition memory
participants presented with trivia statements and asked to the them into a dialogue box and press enter
-1/3 trials: “your entry has been saved”
-1/3 trials: your entry has been saved to the folder Data/ info/ names/…
-1/3 trials: your data has been erased
- then did recognition test -better when asked if it was what they read
Giebl et al (2022)
presented with easy or hard trivia questions
-guessing improved final recall relative to being presented with the answer (pretesting effect). googling right away also produced better cued-recall than immediate presentation, thinking before googling produced best result
landers 92014) theory of gamification
attempts to explain the causal mechanisms by which gamified tasks enhance learning
- make predictions about when and how gamification interventions will enhance learning
- 2 processed: mediating and moderating.
-completed online wiki-based project. gamified group were part of leaderboard
the time on the task significantly produced academic performance
Murayama et al (2010)
Japanese participants completed a stop-watch and watch-stop (control) task. trials randomly intermixed.
- completed 2 sessions, at end of each, participants were given 3 mins to freely engage in either task or do something else
-extrinsic monetary reward reduced number of times participants voluntarily played with stop-watch (intrinsically motivated) task
-undermining effect persisted even when rewards were no longer contingent on performance
Pashler et al (2008)
reviews of carefully controlled studies have revealed no compelling evidence for learning styles. people do not recall ore when information is presented in a way that aligns with their alleged learning style
-focus on strategies that have strong evidence base
Peterson (1991)
students read a textbook chapter - some underlined, some did not.
- 2 months later: MCQ test - some questions probed specific facts, other required inferences by inking different parts of the text
-similar performance on questions that tapped specific knowledge, but the underlining group performed worse than the control group on questions that required inferences
Presley et al (1987)
read series of sentences, elaborative interrogating group were provided explanation software sentences
- in final cued-recall test , the elaborative interrogation group outperformed other groups
