Quiz 5 - Knowledge Acquisition Pt. 2, How to Practice

To what extent does repetitive exposure lead to learning?

Repeated exposure without effort made to structure learning is ineffective, particularly for complex info.

What did Rubin and Kontis (1983) discover using US coins – and what implications does this have?

• People who have generally seen these coins thousands of times still couldn’t remember the details on these coins.

• Maps onto idea that experience doesn’t always lead to expertise, and familiarity does not equal understanding / knowledge.

• Has implications for advertising. Lots of repeated exposure doesn’t mean people will recall.

What was memory champion Feng Wang’s method for learning 300 random digits?

• He made an association for each number from zero to 99 with a particular image

• Mentally constructed a “memory palace” with many specific locations

• Each set of 4 digits can then be encoded as two images. E.g. 63 = banana, 89 = monk. So he remembers monk eating a banana for 6389.

• Sequence of moving through rooms gives you the sequence of numbers.

What is the Method of Loci and how do you implement it?

• Memory palace is a version of method of loci. Involves picturing memory items in specific locations in a set order through a familiar location. Many people use their own home, looking at what is in each room. The route gives you the order of things.

What did Massen et al. (2009) find were the best locations to use in the Method of Loci?

Massen et al. (2009) found a familiar commute to work was more effective for recall than using your home. Potentially due to the order of the route being more well established. You can go around your house in more than one order.

What did Bower (1973) find regarding when the Method of Loci was compared to a no-strategy control?

Presented participants with a string of 100 words. Participants remembered 72% using method of loci, 28% when they did not.

What did De Beni (1997) find when they compared the Method of Loci with rehearsal – for both written and orally-delivered essays?

FILL THIS IN MORE from slides. NOT COMPLETE ANSWER.

Advantage disappeared when passage was read. They argue that reading something is using your visual imagery, so it creates a bottleneck in your brain trying to also use method of loci at the same time.

How does the Pegword system work?

1. Step 1: memorise the ten pegwords (1 is bun, 2 is shoe, three is tree etc.)

2. Imagine the first word on your list interacting with the pegword (e.g. you need to buy a shirt as first item on your shipping list so you remember a bun smeared on your shirt).

How does the Pegword system compare with the Method of Loci (Wang & Thomas, 2000)?

It is equally effective

What are the limitations of the Pegword system?

• You need to do lots of work to remember all the pegwords.

• It’s also harder to use for more abstract memory items, e.g. “ethics” than easy to picture things like “dog.”

Describe the method of remembering names (with your own example) that was demonstrated by Morris et al. (1978) to lead to 80% correct recall.

REWORD:

•Step 1: Search for an imageable substitute for the person’s name. Horswill = Horse + Will (or Whore + Swill).

•Step 2: Select some prominent feature of the person’s face – and link that feature with the name substitute (e.g. HORSe joyfully riding through my eyebrows, having just received a substantial sum from the WILL of a despised relative).

•This method led to 80% recall (Morris et al., 1978).

Under what circumstances did Morris et al. (2005) find that the imagery method of remembering names did not work – and why?

• Under “party conditions,” loud music, background noise. Too cognitively demanding to try and come up with matching pictures.

• Under these conditions, you get better memory by doing repeated retrieval rather than imagery method.

Give your own example of a verbal mnemonic.

Colours of the rainbow: Roll Over You Great Big Ignorant Vegetable

What is the Story Mnemonic method?

A list of unrelated words you have to memorise. You create a story to link all of those words in whatever way, even absurd or surreal.

What are the advantages and disadvantages of the Story Mnemonic method?

• Advantages: doesn’t require previous info to be learned (pegwords where you have to memorise the pegword list or locations require this)

• Disadvantages: takes time and cognitive load to create this story. May not work under constraints on these e.g. party conditions.

Give three reasons why mnemonic memory methods can be very effective (Ericsson, 1988).

1. Encoding is meaningful (you are connecting info to pre-existing knowledge). We know meaning makes things more memorable.

2. You create a retrieval structure that promotes accessibility of items

3. Memory experts practice these techniques a lot, which makes encoding and retrieval both faster

END OF KNOWLEDGE ACQUISITION. NEXT SET IS HOW TO PRACTICE.

What is the most important variable for practice effectiveness, according to Schmidt & Lee (2005)?

Amount of practice.

Describe the difference between “blocked” and “random” practice schedules.

• Blocked practice: do each practice in a big block. Do the same task repeatedly in a big block before moving to the next task. This has low contextual interference as you are not introducing any interference of other tasks between each trial.

• Random practice: You mix up tasks between practice trials. Don’t do the same task over and over. This has high contextual interference. Presence of previous trial will effectively interfere on subsequent trial.

Describe Shea & Morgan’s (1979) transfer design experiment on contextual interference (methods, results, conclusions).

Methods: They had a thing where you had to learn different sequence of knocking paddles with tennis ball as fast as you can. 3 sequences of paddles to be knocked down.

• Random schedule of sequences

• Blocked schedule (keep doing same one of each before moving to next)

• After 10 mins or 10 days they either stay in their condition or swap

Results:

• Blocked practice leads to initially better performance under the learning condition

• BUT random practice leads to overall better long term-learn after retention period and is less dependent on the schedule itself than random practice.

Conclusions: Random practice therefore leads to better learning. Changing things up helps with learning.

BELOW IS ROUGH NOTES, ABOVE IS MARK’S ANSWER.

Methods: Using a tennis ball on device with lights and paddles. Participants had to knock down paddles in a certain order based on the colour of the light that came on. Three sequences to learn. Participants randomly assigned to blocked or random practice schedules.

Results: Initially blocked condition participants were fast. Random condition people start very poorly and catch up a bit. HOWEVER after transfer tests (following 10 minute break), they had 4 conditions (include this in methods). They had random-random, blocked-blocked, random-blocked, and blocked-random. These were the four conditions as some participants were re-assigned after the break. Initially, random people are worse, but when we go to the retention phase (following 10 min break)

THere was also a 10 day break. They were reassigned over the 10 day break.

GO TO THIS PART OF THE PRESENTATION to do this flashcard.

Conclusions: Could show that people in blocked group are learning more effectively than those in the random grou, BUT there is a confound. Could be that random group has learned just as well, but there is a performance deficit being forced on them because their job is harder. Constant task switching could lead to a temporary performance deficit. So we can’t tell from this study if these blocked vs. random practice effects are lasting. This same issueoccurs on other training study designs.

What’s a transfer design?

Helps resolve the confound in Shea & Morgan’s study.

Give people a rest then test their performance again. After this break, now the blocked practice participants perform worse than

Name the design that Karpicke & Roediger (2008) used in their 4 condition experiment on the testing effect.

Copy over answer from previous week.

What are the two benefits of using a random rather than a blocked practice schedule?

REWORD

•Blocked practice produces better initial performance than random practice.

•However – random practice leads to better and longer term learning than blocked practice (i.e., performance during acquisition does not necessarily reflect learning).

•Also – random practice leads to less context dependency. That is, participants who used a random practice schedule could do both blocked and random tests – but participants who did blocked practice were only good at the blocked test.

What is contextual interference?

REWORD:

•“Contextual interference” is the learning benefit you get from using a random practice schedule rather than a blocked practice schedule (Broadbent et al., 2017).

How did Van Merrienboer & Sweller (2005) reconcile contextual interference effects with Cognitive Load Theory?

• Contextual interference you would think would lead to less effective learning as it contrasts Cognitive Load Theory that additional cognitive load would overload cognition, leading to worse learning.

• However in contextual interference, the extra difficulty you’ve introduced is useful for the task. It’s relevant to learning the skill.

• However in cognitive load theory, the stuff interfering is irrelevant. It’s not part of what you’re trying to learn. It’s irrelevant cognitive load that is unhelpful in Cognitive Load Theory. The theories do align. Not all additional cognitive load is bad for learning.

Describe a contextual interference experiment involving a simulation of a real sport.

REWORD:

- Broadbent et al. (2017) asked novice tennis players to predict where a ball would go, while watching live size video of tennis players hitting three types of tennis shot (type of shot presented as either random or blocked).

• Clips cut to black when the racket contacted the ball. Participants called out where they thought the ball would go (4 possible locations). For feedback, participants saw the clip again after each shot, this time showing where the ball went.

• No significant difference in accuracy in pre-test. The random condition led to better performance in the delayed (retention) test than blocked. Random condition had learned more, even though their performance was suppressed during learning because a random schedule is harder.

Describe three theories that explain the contextual interference effect.

These theories aren’t mutually exclusive. Could all be true.

REWORD:

1. Elaborative Processing Hypothesis

   1. Shea & Morgan (1979) argued a random schedule forces people to engage in more elaborative & distinctive conceptual processing. People learn from comparing and contrasting the different items.

2. Action Plan Reconstruction Hypothesis

   1. Lee & Magill (1983) argued that participants have to reconstruct how to do a task when it is preceded by a different task. The increased reconstruction/remembering/retrieval in a random schedule promotes memory processing for learning (i.e. it’s the testing effect). Basically, they have to do lots more retrieval, so via testing effect, leads to better memory.

3. Error Processing Hypothesis

   1. Random schedules lead to more errors. Errors (with corrective feedback) promote learning (Broadbent et al., 2017).

What evidence suggests that trainees might be unlikely to choose a random practice schedule if given a choice?

REWORD:

Simon & Bjork (2001) conducted an experiment in which participants learned 3 keystroke patterns on a number pad using either a random or blocked schedule.

They did a retention test 24 hours later.

The random group performed better BUT predicted their performance would be worse. Just like the testing effect. People feel they’re learning more under the blocked strategy, but they’re actually learning more under random strategy.

Make a case for the generalizability of the contextual interference effect.

• They found it works when you interleave maths problems rather than presenting them in blocks.

• Also for painting. Students learning styles of different artists did better when presented in random order than blocks. Against, students believed blocked would be more effective.

• These contextual interference findings were well replicated for some laboratory tasks, different sports.

Describe Bourne & Archer’s (1956) experiment on massed vs distributed practice using a pursuit rotor tracking task (methods, results, conclusions).

This is about how you spread out your practice time.

Massed practice: doing study/practice periods together without breaks

Distributed practice: breaking up work/study periods with rest or another task

Methods: (reword) used a pursuit rotor tracking task and 5 different practice distributions (0 rest to 60 secs rest between 30 sec trials).

Different groups were: 60 second rests, 45 second rest, 30s, 15s, no rest. Outcome was percentage time on target. Then on transfer trial where everyone had a break, longer rests groups still did better order was maintained.

Results: People with breaks did better. Clear descending order of more breaks = more time on target in motor tracking task (moving stylus around moving object following targets). Longer breaks led to better learning after transfer trial, not just about being overloaded during tasks explaining less rest meaning worse performance.

Describe Baddeley & Longman’s (1978) experiment on massed vs distributed practice involving postmen (methods, results, conclusions).

REWORD:

•The British Post Office introduced post codes in the 1970s, which required postmen to learn to type the codes into a keypad. Baddeley & Longman (1978) ran an experiment to determine the most effective way to schedule their practice.

•Postmen were randomly assigned to 4 conditions:

–2 x 2-hour sessions per day

–1 x 2-hour session per day

–2 x 1-hour sessions per day

–1 x 1-hour session per day

Results: •The most distributed group (1 x 1-hour) learned faster, continued to improve at a faster rate, and had better skill retention after several months compared with most massed group. However, this group were also least satisfied with their schedule and the overall duration of their training was longer (because it was more spread out). People want to do massed practice over distributed (people want to get it over and done with).

Describe Moulton et al.’s (2006) experiment on massed vs distributed practice involving surgeons (methods, results, conclusions).

REWORD:

•Moulton et al (2006) assigned surgical residents to either receive skill training (microvascular anastomosis) in a 1-day session (massed) or 4 shorter sessions spread over several weeks (distributed).

•Both groups performed a battery of skill tests (including performing the skill on a live rat) to the same level immediately after training.

•However, the distributed group were significantly better in these tests 1 month after training.

Conclusion: Another bit of evidence for distributed practice over massed practice. Distributed better than massed.

Give 3 reasons why distributed practice might lead to better retention than massed practice.

REWORD:

1.The breaks give the brain a chance to better consolidate the new skill or the information learned (e.g. Moulton et al 2006).

2.Distributed practice allows for more cognitive preparation and mental rehearsal between sessions (where this could be implicit).

3.Also distributed practice may increase the amount of retrieval required, just like random practice schedules (boosting consolidation and strengthening retrieval structures).

What’s the spacing effect?

•The distributed practice effect is sometimes known as the spacing effect (e.g. Pashler et al. 2007). EXPLAIN WHAT DISTRIBUTED PRACTICE EFFECT IS.

What was Ebbinghaus’s (1913) “memory savings” discovery?

• Ebbinghaus noticed that when we forget things it’s not the same as losing that info from our brains entirely. Not being able to retrieve is not the same as the info being deleted from our brain.

• He memorised a list of words until he could remember them all perfectly.

• He waited until he forgot the entire list.

• He then tried to re-memorise the list.

• He found memorising it the second time was much faster the second time. This shows the information was still there, you’re just reactivating the pathways to recalling it rather than starting from scratch. This is memory savings.

Bahrick & Hall (1991) found people who didn’t forget algebra over 55 years. What was their secret?

REWORD:

•Bahrick & Hall (1991) found evidence that this “reloading” effect can eventually immunize against forgetting.

•Students who took one algebra course, forgot their algebra over time. However, students who had later exposure to algebra (e.g. in later Calculus courses) became immune to forgetting.

•They were just as good at algebra after 55 years as when they had just been taught it.

•Willingham (2021) argued that this was unlikely to be due to students who complete more education being less prone to forgetting (which would be a potential confound) because higher performing students (more likely to do more education) have been found to forget at the same rate as lower performing students (Ellis et al., 1998).

So the secret is retrieval. They retrieved it as part of a later calculus course.

What benefits does sleeping have for learning and why?

• No evidence we can learn while sleeping (like listening to audio while we sleep)

• HOWEVER there is lots of evidence that sleeping helps us consolidate information we have just learned while awake

  • Synaptic consolidation: to do with hippocampus, works in about 24hrs

  • Systems consolidation: slower, info transfer from hippocampus to other parts of brain

  • Basically, your brain is consolidating info

• Sleep benefits for learning are established to work long after initial learning

Describe Mazza et al.’s (2016) experiment demonstrating how sleep can help learning (methods, results, conclusions).

REWORD:

•Mazza et al. (2016) made participants learn the meanings of 16 Swahili words until they got them all correct.

•12 hours later, they made them learn the words again. So two sets of learning.

–Half the participants did not sleep between the two sessions (9am vs 9pm).

–Half the participants slept overnight between the sessions (9pm vs 9am).

•Sleeping between sessions halved the number of trials required to relearn all the words in the second session.

•They also found that sleeping led to better retention for the words both 1 week and 6 months later.

Describe the different types of variability that may affect the performance of motor skills.

1. Execution variables: like how hard you throw, where you aim. The actions you’re doing.

2. External variables: what other people around you are doing, the environment you’re in, your stress levels, if in front of an audience

What is schema theory?

• People learn schemas (particular rules) to help them perform an action. A script in our head for the action.

• Each action in a task (e.g. passing vs. kicking a ball) is a different class of action. The schemas are like motor programs in our head.

• REWORD:

  • •A schema might take the form: “going through this particular set of motions” results in “me successfully throwing the ball to my team mate”.

    •Learning the action is about refining the parameters of the schema (e.g., how much force you put into your throw) to produce the various specific outcomes you want (e.g., ball travels to a team-mate 3 metres away, slow enough for them to catch it but fast enough to avoid interception).

What did Schmidt (1975) suggest would be beneficial for ensuring the development of a generalizable schema when learning a motor skill?

We want to intentionally vary how we practice in as many ways as possible in line with how these dimensions may vary while we perform the skill.

E.g. performing on different surfaces of tennis courts, different times of day, different weather.

Describe Catalano & Kleiner’s (1984) experiment investigating the effects of variable practice (methods, results, conclusions).

REWORD:

•For example, Catalano & Kleiner (1984) asked participants to predict when a moving light would arrive by moving their hand (i.e. analogous to catching a ball). They have to move their hand to intercept it as if it’s a ball they’re trying to catch.

•Some practiced this task under constant conditions: 5, 7, 9, OR 11 mph. They practice in only one of these speeds.

•Other practiced under variable conditions: 5, 7, 9, AND 11 mph. (So they varied how fast the light moved).

•When they were all given a transfer task that involved new speeds OUTSIDE the range of the practice speeds (1, 3, 13, and 15 mph), the variable group performed much better. ((So no one practiced under these speeds, but the variable group who were used to swapping, more variability in their practice, they did much better even though they hadn’t practiced under these particular speeds. It’s not about matching the speeds. Schmidt would say that variability has made that schema for that task much more robust to variability, even variability you haven’t experienced before.))

•This supported Schmidt’s variable practice proposal.