Hello All,
Next week we have 2 articles to read. The Engle article is an overview of the viewpoint of Working Memory Capacity as Executive Attention. We will not directly dicuss this article, but it does give a brief/clear conception that should help interpret Jarosz's findings. We will primarily discuss the Jarosz article. Please write a discussion board post by 9PM Monday, November 14th. Daniel will lead this discussion.
See you at 3PM on Wednesday, November 16th. There will be pizza since this is our last class for the semester!
Hope you all have a nice weekend.
Best,
Dr. Braasch
I really appreciated the Engle article. It did a great job explaining what working memory is and giving examples of how it can be measured. I didn’t know that working memory is different from short-term memory, nor did I know that it is really more like executive attention rather than actual memory. I thought the three examples were interesting to read about, especially the antisaccade task. I would have guessed that people who have better control of their attention would perform better, but I was surprised that there was such a large difference between the high and low working memory groups. In terms of conceptual replication, it seems like all 3 tests described are different ways of measuring the same construct, working memory. I’m interested in learning about other ways that it can be measured! I do have one question though: what does it mean when the author says that working memory is domain-free? I think I have a vague idea but I’d like to hear what other people have to say.
ReplyDeleteI enjoyed reading the Jarosz & Wiley article too. I think IQ tests are kind of fun to take, so it was interesting to learn about how working memory capacity and the use of different strategies impacts performance. I was really impressed that researchers extended the first experiment to include the use of an eye tracker, because it’s not a method I would think to use to analyze how a person picks an answer. It’s a great example of conceptual replication. It made me wonder how else this research could be replicated, and I’m not a cognitive psych person, so I would like to hear ideas that other people have!!
I personally am very interested in memory, so this was an interesting set of articles for me.
ReplyDeleteWhen referring to people as having a low or high working memory capacity, I am curious as to how those are defined? I also wonder if this is something considered to be inherent or if we can train people to have a higher working memory capacity in the same way we teach people to ride a bike. Does practice make a difference? I find it hard to believe that WMC is a stable construct, although I am sure a lot of it has to do with individual differences, similar to most cognitive constructs.
When talking about WMC, I also wonder how it is affected by relevance. Similar to the spacing effect we talked about yesterday, I would assume you can push your WMC to a higher capacity if you know you will be tested on that material within a short timeframe after memorization/rehearsal. I also would assume WMC can be affected by simply interest. I'm probably more likely to remember a list of concepts/words that are interesting to me than I am to remember a list of words that have no meaning to me. These are a few topics that I wish the articles had addressed.
I also wonder if lower WMC is related in any way to certain disorders, such as ADHD or ADD. I would assume so but I do not have much background knowledge in the issues behind these disorders.
Looking forward to our last class meeting! (and pizza)
The Jarosz & WIley paper was an interesting read regarding the interplay between general fluid intelligence and working memory capacity. Part of me, when I first started the paper, had a reaction of “Umm… of course.” It seemed a bit obvious that people who are strong in one cognitive capacity would be strong in another. A rising tide raises all boats situation. But as the paper progressed, I realized that it was focusing on the Why? or How? Part of the situation.
ReplyDeleteI liked how they went into a description of multiple possible explanations before explaining why theirs, the attentional control account, was probably the best. Their first experiment excellently showed took the most distracting standard answer of the RAPM and either removed it or kept it in. They said that this was making the most salient/distracting incorrect answer. I’m not sure I agree with their methods here, but I can’t think of any way to improve it. I don’t know enough about the area to know if there are different ways of increasing salience. I think they assume correlation implies causation with this method and while that’s not good, it might be the best possible. I think another way might be to use external distractors to test the ACA. It might better show that people who can tune out, say distracting headphone tones, than others tend to perform better. That would be an interesting replication of their first experiment: seeing if something that is definitely, unquestionably a distractor yields similar results.
Their second experiment was also interesting. It was designed to find a lurking variable in their first experiment, which it did. I like their explanation, which is that higher WMC people go into the answer bank with an idea of the answer, while lower WMC people search for an answer that looks right.
I suppose all in all this is a good study that I think could have used some more stringent methods. Testing the ACA might be better done by examining outside competitors for attention, like a persistent vacuum going outside, rather than inside, their high salience wrong answer.
The Engle paper was also a good read. I appreciated their multiple tests to confirm their hypothesis. I have done some work with WM before, but their distinction between it and STM is very well made.
Again really enjoyed this week’s reading. The Engle article was a great review of working memory capacity, I appreciated the article’s comprehensive overview WMC relationship with proactive interference as well. The Jorosz article was so interesting, I personally enjoy memory and more specially working memory. I especially appreciated the simple reminder at the beginning of the introduction that we are still (after decades and decades of research) are trying to answer the question “What is intelligence”?
ReplyDeleteWith respect to the RAPM task, I am so impressed with its complexity and how it was modified for this study. I think modifying existing inventories and tests are favorable in comparison of creating new ones when dealing with replication. That’s just personal opinion, but the adjustments made for the present research were clever and revealed quite a bit about WMC. I think the results of experiment 1 and 2 were expected but very revealing about how low and high WMC individuals handle salient distractors when problem solving. I would like to know more about the mechanisms that allow high WMC individuals to better avoid distraction from the highly salient incorrect options when taking the RAPM.
I very much admire anyone who works with eye tracking and analyzing eye tracking data, as it appears so complex. I think using it as a way to expand on experiment 1 was not only intelligent but shed a lot of light on how different problem solving strategies vary in susceptibility to salient distractors. In the final discussion, it is made clear that it is not just susceptibility to distraction, rather a variation in strategy use may intensify WMC performance on the RAPM items. This is interesting to me; I feel like this is pretty obvious but what cognitive mechanisms are directly in play here? Since the authors are brave enough to tackle eye tracking, perhaps something like employing a “think aloud” as participants complete the RAPM could further reveal relationships between WMC, problem solving, distracters, etc.
While I am quite familiar with working memory, I think the Engle article did a great job at introducing the idea of it being a concept closely related to executive attention, rather than just being solely associated with memory. This is what makes the difference between working memory and short-term memory.
ReplyDeleteAs far as the Jarosz study, I enjoyed this as well. It definitely took an interesting route regarding how to measure and relate the ideas of working memory capacity, bringing into play the asset of attentional control, and general fluid intelligence. However, the low sample size in the second experiment, while confirming some of the results of the first experiment, warrant some caution. I fear that some claims for the second experiment, despite significance, may not be fully justified because low sample sizes such as these can make regression equations unreliable. I would like to see if similar results for these models occur in a larger sample size. I’m also unsure how I feel about “marginal” significance, though I do believe it noteworthy.
To take the theme of replication into view, I have one outstanding question in my mind. As far as the second experiment, the claim is made that the differing solution strategies are influenced by working memory capacity. These strategies are adjustable; they can be taught to individuals. I am personally curious, if you manipulate solution strategy (i.e. train people to either look at the problem, think about it, decide a solution, then look at answers or to look at the solutions and see which fits best), if a similar effect would be seen.
The Engle article is very helpful for me to understand working memory system and what general fluid intelligence is. Because I am taking structural equation modeling course this semester, so it is very interesting to see another application of this analysis rather than just testing the relations between constructs in the model. The way the author and his colleague use structural equation modeling analysis to test whether they should combine two highly correlated constructs (WM and STM) by comparing the model fit is very convincing as a statistic approach to separate WM and STM.
ReplyDeleteThe review of Engle article did a great job on paving the way for understanding Jarosz and Wiley’s article. Both experiments are very interesting, the theoretical grounding is sound, and their findings are consistent with the previous research. However, I was concerning about the sample size and whether the participants can stand for the population, especially for the second experiment. Another concern is that since the authors use hierarchical regression, but they did not mention whether they met the assumptions such as normality of residuals and mulicollinarity. So the results would be more trustworthy if they can provide such information.
In terms of replication, I am curious to see how working memory capacity and problem solving strategies relate to academic tasks. The two primary strategies are for solving visual analogy problems, but I remembered I used both strategies on multiple-choice questions in the exams. For example, if I know what the answer is, I will use constructing matching, and if I am not sure the answer, I will use response elimination. So it would be interesting for me to see the relationship between the strategies and working memory relate to academic tasks.
Can someone save me some pizza?
ReplyDeleteIn the Jarosz article I thought it was clever that they thought the multiple choices were acting as distractors. I don't think I would have thought of that. An unintended consequence of reducing the number of choices from 8 to 4 though is that they have a much better chance of selecting the right answer. Looking at their results I don't think this was an issue though.
I always enjoy reading about eye tracking studies. Since the eye fixations was negatively correlated with WMC, it suggests to me that increasing the number of choices (distractors) wouldn't have much of an impact if you already know the answer. This goes back to all the experiments on the visual world paradigm, for example, if you flash 4 pictures on the screen at the same time and say a word, people stare at the picture most relevant to that word without much delay at all and then fixate on it and the other pictures only get a little attention at the start.
This is really good paper though, I like it. It is hard to say much when you can't find any big potential flaws.
I thought the Engle article was useful in giving a general overview of working memory. I have not spent a lot of time with working memory research, so I thought this article was hugely helpful to me. It certainly laid the groundwork that helped me understand the Jarosz and Wiley 2012 article. From Jarosz and Wiley article, I found the concept of attentional control account to be very interesting.
ReplyDeleteFrom the second experiment, I did have an issue with the low N. However, I couldn't find any huge potential errors with it overall. I would like to see whether or not these results will replicate if participants are not given the answers right away. If participants come up with a solution and then are presented answer options, will the salient distractor items be as distracting?
I’ve always been a fan of reading up on and exploring different factors that may interact with working memory capacity. One such factor that I have had some experience with in my work is the idea of attentional control and what factors may play role in interacting with it. My previous research has also investigated visual saliency and how visually salient items can “steal” activation away when viewing an image as saliency tends to be simply a winner take all (WTA) phenomenon. To see Jarosz apply saliency to WM tasks was an interesting take and one that I had never considered outside of the realm of typical visual cognition (visual world paradigms, blank visual world paradigms, etc.).
ReplyDeleteOf the two studies, I want to devote some space here to talking about Study 2 as this is where the eye tracking action was! In my thesis work on misinformation, I found a similar result in that participants seemed to be actively suppressing incorrect information during retrieval and not devoting a very high proportion of there fixations to incorrect answer selections even if they had been previously primed to do so. I hypothesized that WMC most likely played some type of role in participants ability to suppress salient lures and it looks like Jarosz found a similar effect which is awesome!
I also found the finding of less toggling = greater suppression intriguing in that it definitely stands to reason that if you are successfully suppressing a salient competitor, then you should not feel the need to toggle as often as you have more or less eliminated that possible answer selection from the board and taken it out of play. All in all I really enjoyed this paper—even if it only selfishly helped me confirm a few theories about my own research ;)
The Jarosz article seemed to be relevant to both discussions of replication and application. In terms of replication, the article provided another unique demonstration of how researchers might form hypotheses to better guide attempts at replication. Once again, as we have seen in some of the past weeks' articles, getting general relationships to replicate is relatively easy, at least if they are broadly defined, while getting explanatory mechanisms underlying those relationships to replicate is a much more difficult endeavor. It was insightful to see how the researchers manipulated their research questions and predictions to shift the focus from the "learning account" to the "attentional control account". It would seem that success in more specific replication lies, in part, with how well one is able to effectively make these shifts without leaving too much in to confound alternate hypotheses.
ReplyDeleteIn terms of application, I found the discussion that better problem solvers are often those that use more constructive means of solving problems to be both interesting and somewhat surprising. While it is not surprising that being able to come up with a suspected correct answer prior to looking at the options is beneficial, what is surprising is how often this notion conflicts with common test-taking strategies and behavior. Though students are always encouraged to work through questions first, there are still a number of test-taking strategies that highlight looking through the answers rather early. In these instances, one often strikes through known distractors and then returns to the question to seek which information best fits with the remaining answers. While this example may relate more to differences in the type of intelligence being tested, crystallized vs fluid, etc., I wonder how regularly it applies across different aspects of intelligence measurement, particularly when instrument sub-scales frequently tap into different factors of intelligence simultaneously. I also wonder if this tendency to look at the response bank earlier in problem solving may be influenced by the length of time allotted for the problem solvers. For instance, while I may go through the steps of constructing my own anticipated answer if I am given plenty of time, I am much more likely to quickly jump to the given answers if I am in a strictly timed situation, even if this is ultimately less effective and more time-consuming. Again, whether this is an important factor depends on whether one is measuring achievement or ability and on the specific domains being targeted. Further, I would anticipate lower WM problem solvers to suffer more from restricted time, given its general positive correlation with a range of successful aspects of problem solving, but higher WM students may also be significantly impacted in these scenarios, if only to a smaller extent.
Ending with a small aside, I wonder what would happen if one were to place a mandatory hold on the presentation of the response bank. That is, regardless of one's WM categorization or speed working through the items, if one were given no option but to first work on the problem before seeing the answers, I wonder if the gap between lower WM and higher WM solvers might diminish. Still, this is rather unlikely given that the lower WM individuals might have concurrent deficits in other resources compared to higher WM individuals.
I can’t believe this will be the last class! It was a very good class for me with so many different topic of cognitive science. Thank you Dr Jason and the class.
ReplyDeleteThis week reading is very interesting. The article mainly talk about WMC reflects the ability to control the focus of attention and resist interference and distraction. The hypothesis was examined by manipulating the level of distraction among the incorrect responses contained in RAPM problems by varying whether the response bank included the most commonly selected incorrect response. The second experiment used eye tracking to demonstrate that these effects seem to be rooted in differences in susceptibility to distraction as well as strategy differences between high and low WMC individuals. Results are discussed in terms of current theories about the role of attentional control in performance on general fluid intelligence tasks.
In the first experiment the hypothesis by manipulating the presence of the most frequently chosen incorrect responses within the response bank of RAPM items. In addition to manipulating the presence of highly salient incorrect responses, one can also consider qualitative differences among error types. Finally, an attentional control account, like the learning account, makes no predictions regarding capacity, and as such would not predict that the selection of incomplete solution responses would relate to WMC. On the other hand, an attentional control account predicts that low WMC individuals would be more likely to be affected by proactive interference.
In the second experiment the regressions were used for analyses despite the low number of participants in an effort to provide results in as parallel a manner as possible to Experiment 1. Results remain similar using median splits and an analysis of variance approach. Experiment 1 in demonstrating that performance on high salience items predicts WMC over and above performance on low salience items. This is because strong effect in the high-salience conditions. these data suggest that low WMC participants are more susceptible to distraction from highly salient incorrect responses.
Overall results of the two experiments make several suggestions about the relation of WMC to RAPM performance. Some of the findings make sense in the greater literature on visual analogy problems and strategy use. Study extends early findings on differences between more and less successful problem solvers to demonstrate that one factor influencing the selection of a particular solution strategy is the WMC of the individual.
And finally I see why the author brought two experiment together. And the author this nice job doing that.
This semester has totally flown by—We’ve definitely covered a lot of ground, and have looked at so many facets of replication thus far. I thought that this last topic was a nice wrap up, because it was a perfect representation of using a conceptual replication within the same study.
ReplyDeleteEssentially, the study’s purpose was to show that working memory capacity is explained more by a model of attentional control, whereby they used a modified version of the RAPM to assess for the types of errors made people with high WMC and low WMC, and also references eye movement trackers to affirm the types of attentional strategies that were used by the two types.
I thought the study was well designed in terms of its methodology, however there were a few aspects that I’m a bit critical of. My first concern relates to issues in sampling—All of the participants were University of Chicago undergraduates in their subject pool, who are likely to already have higher intelligences, given the prestige of the university. The samples themselves also seemed rather small, especially for experiment 2. This reminds me of the problem of generalizability that we had with the Boroditsky article. It also was a little bit unclear to me how they divided the groups into high and low WMC? I’m curious about how this study would do if it were directly replicated in another university, which would have perhaps a greater range of ability.
Another thing that I was curious about was this emphasis on “strategy.” Could it be possible to “coach” people to have a certain strategy? Or maybe there’s an element of their educational background that emphasized certain cognitive mindsets that would make them more likely to use these strategies? I’m curious about the metacognitive side of the test—What exactly was going through their heads, and what their thought process on things was. It might be interesting to include a questionnaire asking about these elements.
The Engle article provided a solid structure for understanding Working Memory. It reveals an interesting line of questioning regarding Working Memory capacity and executive attention.
ReplyDeleteI found myself most impressed with the dual experiment design employed by Jarosz and Wiley. Analyzing eye movements in a second experiment to get at strategy differences was a great attempt to get at some more underlying information. Their findings are certainly enlightening.
I find it particularly interesting to consider their authors' use of new RAPM problems and the differences in performance from the original. While the authors have provided justifications that their new problems should maintain validity, any deviations are something to always keep under careful consideration.
Although I do like think and learn the concept of working memory, since working memory is the crucial step to consolidate perceptual information into solid knowledge, most of knowledge about working memory of mine is not far from the lecture that I was given in Memory class. this Jarosz (2012) article’s idea and experiments are quite unfamiliar to me, yet, it was still interesting.
ReplyDeleteIn this article, they tried to figure out whether working memory capacity (WMC) and Raven’s Advanced Progress Matrics (RAPM), which is used to investigate relationship between working memory capacity and general fluid intelligence, are related. In this study, authors took attentional control account which demonstrates that high WMC individuals are better able to resist interference from previously encountered items and generate new solutions to new items. Moreover, this account is consistent with the recent result that WMC could be related to attentional control, or the ability to control focus of attention. Therefore, in this research, authors tried to look into whether distraction could be the another factor affecting the RAMP’s correlation with WMC. Both experiment 1 and 2 showed that high salience items predicts WMC.
In terms of replication, since experiment 2 of this research already replicated effect of experiment 1 so I think it could be interesting to see whether this effect is observed in neurophysiological methods rather than behavioral methods. Since brain waves are quite different when we focus on something or distracted, it would be interesting to investigate this attentional control in terms of relationship between WMC and general intelligence(RAPM).
I enjoyed how the Jarosz and Wiley article was written. Unlike last week’s reading, I felt like this article clearly explained everything, and the way it was organized seemed to flow more logically. The RAPM reminds me of a task on the WAIS IQ test that I took recently to help a fellow student practice for their class, so I was very interested in reading some research about tasks like this.
ReplyDeleteWhen I first started to read the article and the authors had mentioned how performance on the RAPM is related to working memory capacity, but that it was not completely clear why this was, I immediately thought of the occurrence of proactive interference (specifically how people with high working memory capacity are less susceptible to it) as a potential reason for this relationship. I have done some research on proactive interference and working memory capacity, so I was excited to see this mentioned as a possible explanation in the attentional control account (and to ultimately have it be a solid explanation).
I was also glad to see that the authors did not exclude participants by only retaining the upper and lower quartiles of participants in the working memory tasks. I am always a little uncomfortable with removing the participants who are not at either extreme because it gets rid of so much of the variability that exists within the population where people are all over the working memory capacity distribution.