Friday, September 25, 2015

Reading Speed and Memory for Prose

Dee-Lucas, Diana (1979)

Summary:
Through this study, the author investigated the effects of reading rates on the retention of the information being read. Using two groups of twenty undergraduate students, a fast group and a slow group, Dee-Lucas developed an experiment where the subjects read three passages of approximately 415 words on the topics of Hinduism, typhus, and the history of Alaska. After reading the passages, the students were asked a set of eight questions, four fill-in the blank questions and 4 short essay questions. The questions tested a range of knowledge from verbatim recall to memory of more general ideas. Students in the fast group were paid a varying rate based on how quickly they read the passage (under 60 seconds, 60-90 seconds, 90-120 seconds, etc., up to 180 seconds) and how many questions they answered correctly. Students in the slow group were paid a flat rate based on how many questions they answered correctly with minor incentives for reading faster (under 200 seconds or under 180 seconds).

The author developed scores to test the type of information remembered, where each proposition, or item of recall, was assigned a “recallability score” based on the number of slow readers who recalled that proposition, and a “susceptibility to loss” score, which was the proportion of recall loss suffered by each proposition when students read faster. The propositions were broken down into case propositions (those that show relations between objects and maintain continuity of the passage), descriptive propositions (which express details and attributive relations), and those which express relations of other types. The author also tested recall accuracy, assigned as “identical,” “gist,” or “incorrect.”

Results:
The payoff structure was effective; the mean reading time for the fast group was 99.3 seconds and 207.8 seconds for the slow group. Retention, however, suffered with speed, as mean recall was 30.3 out of 129 propositions for the fast group and 47.3 out of 129 propositions for the slow group. The case propositions, which state the action of the passage, made up a majority of the recalled propositions in the fast group, while attributive propositions, which expressed descriptive information, made up most of the set of poorly recalled propositions. In terms of accuracy, there were no significant differences between the fast and slow groups. Therefore, while the fast group retained less information overall, the information they did retain was as accurate as the information retained by the slow group. However, fast readers may have missed implicit causal relationships that were picked up by the slow readers.

Critique:
This article basically states that when individuals increase their reading speed, they risk losing some of the information in the passage; however, the information lost is primarily descriptive information as opposed to continuity information. This could potentially be detrimental to intelligence analysts where the details are frequently just as important as the overall narrative of a piece of literature. Additionally, the faster readers missed implicit causal relationships; while some information will be explicitly expressed, many of the answers that we as intelligence analysts seek require a deeper examination into the message that is left unsaid or the intentions behind an action. Finally, a flaw in this study is the lack of an assessment of the difficulty of information being read and recalled. The fact that the slow group only recalled 47.3 out of 129 propositions (37%) suggests that the information may have been above the subjects’ level of comprehension, resulting in difficulty recalling the information.

Source:
Dee-Lucas, D. (1979). Reading speed and memory for prose. Journal of Reading Behavior, 11(3), 221-233. Retrieved from: http://jlr.sagepub.com/content/11/3/221.full.pdf+html 

16 comments:

  1. I think this is a decent study in order to show the difference between two kinds of people who have nearly the same capabilities: one who tries to read fast to get more money and the other who reads comfortably and will get a flat rate money. However, the study falls short in showing difference between one who knows how to read fast effectively by applying speed reading techniques and the other one who doesn't have the speed reading ability.

    ReplyDelete
    Replies
    1. Ertugrul, I agree completely. The authors do not provide any training or instruction to the fast group, which would likely increase their comprehension and recall of the material.

      Delete
  2. This comment has been removed by a blog administrator.

    ReplyDelete
  3. I believe the mental capabilities of the contributors also should be taken into consideration in this experiment. Because, it may directly affect the experiment results along with the reading speeds of the contributors.

    ReplyDelete
  4. Andrew, I completely agree with your critique regarding speed reading and the possible detrimental effects for intelligence analyst. With that being said, do you think there are any situations where speed reading is beneficial for analysts? Or do you think the costs always outweigh the benefits?

    ReplyDelete
    Replies
    1. Oleg, one of the big issues with this study that Ertugrul pointed out above is that the fast group has not received any training or instruction on speed reading, training which would likely improve comprehension and recall. I'm sure speed reading can be beneficial for analysts, as Prof. Wheaton mentioned in class, if you can improve the efficiency of your analysts, you're basically getting another analysts without paying for one. So I think training and practice in speed reading would be beneficial to analysts, whereas simply being told to read faster is detrimental to their comprehension.

      Delete
  5. Andrew, I think speed reading is beneficial for analysts, especially in the early phases of research when an individual may not know what to look for when it comes to terminology and/or subject matter comprehension. I do, however, believe an analyst should go back and reengage with said information to solidify concepts.

    ReplyDelete
    Replies
    1. I agree with you, when you're just trying to get up to speed on a subject, often quantity over quality is an appropriate strategy. The more you read, particularly on a specific topic, the better understanding you'll have of the topic overall. But when needing facts, such as population or military capabilities, or for techniques such as ACH, slowing down to ensure full comprehension of the details will be more beneficial.

      Delete
  6. I thought this was a very interesting article overall and I agree with your critique of it. I found it especially interesting that the difference in the amount of propositions that were remembered was not very big between the slow group and the fast group. I also agree with Ertugrul's comment about no application of a speed reading technique and am a bit disappointed that they did not do that. I feel that the results would have been much more significant had that been the case.

    ReplyDelete
    Replies
    1. I agree with your last statement, and the author does note that further investigation is needed on the subject. With the application of speed reading training and techniques, I would not be surprised if the fast group actually surpassed the slow group in terms of recalled propositions.

      Delete
  7. Based on this article it seems like a choice must be made between speed and memory, but they are not proportional to each other. The slow group roughly doubled the time it too to read but only answered half again as many questions right. I think the overall numbers show that speed reading can be a useful tool, but needs to be used with caution in intelligence since so much can be missed.

    ReplyDelete
    Replies
    1. Dan, I agree with your statement, but I think the fast group's performance could have been dramatically improved through speed reading training instead of just being told to read quickly. Additionally, the author does note that this experiment focuses on the importance of being able to alter reading speeds based on the text as opposed to forcing a choice between speed or recollection. The two options function more as the extremes of a scale on which you, as the reader, move back and forth based on the situation.

      Delete
  8. The only weakness I see is the "fast group" doesn't appear to have gotten any training on speed reading. I think it's unfair to judge t

    ReplyDelete
    Replies
    1. **unfair to judge the results without speed reading training first.

      Delete
  9. Reading everyone's comments made me think about the fox vs. hedgehog argument Professor Wheaton brought up in class. If speed reading enables you to know more little things (just like the foxes) and less of one big subject or article (like the hedgehogs), then could you argue that analysts who speed read make better forecasters?

    ReplyDelete