Talk:Issues in Interdisciplinarity 2020-21/Evidence in Intelligence Testing

Log of Meetings
20.11.20 (Natasha, Benjamin)

We came up with a working title relating to AI. Brainstormed other ideas, and scheduled another meeting. Created a google docs folder.

22.11.20 (Natasha, Benjamin, Charlotte)

Talked more about our different ideas and added them to the google doc, decided to come to a final decision of the topic by the end of the week at the latest.

23.11.20 (Natasha, Benjamin, Charlotte)

Finalised our topic, talked about the structure and different sections. Assigned out the sections to the group.

30.11.20 (Natasha, Benjamin, Charley)

Exchanged findings and discussed potential areas of conflict between disciplines. Agreed on areas to focus on and which to forget. Each to continue researching topic and submit initial contributions this week. Next two weeks to be process of finalising introductions and conclusions.

06.12.20 (Natasha, Benjamin, Charley)

Online meeting to discuss how the project is going. We focused on the interdisciplinary conflict, how to link the topics together, and making sure we are completing the task in the way that meets the marking criteria. We are going to continue our individual sections and begin integrating the topics collaboratively through the intro/conclusion before finalising next week.

Measuring Intelligence in the Social Sciences Entry
I have just written a brief outline. Will return to add references and make changes, but just to give you both an idea of what this section will be saying so we have time to discuss and make all the necessary adjustments Xrcaatnp (discuss • contribs) 19:13, 2 December 2020 (UTC)

I have restructured this section and labelled it 'An Anthropological and Sociological Perspective' because I think the two offer a distinct way of looking at the issue of measuring intelligence which differs quite a lot from the others. I have rewritten it to make it more clear the point I am trying to make about these approaches and how their different take on it produces conflict, but will no doubt have to make some more changes again so that we can link it with the issue of computers. If you have ideas on how to do this feel free to make changes! (Will return to add references again). Xrcaatnp (discuss • contribs) 12:37, 5 December 2020 (UTC)

Measuring Intelligence in Computer Science Entry
I have added the topics for this section that I will be elaborating on over the weekend for reference incase they spark any other ideas/conflict ideas for you both. Camjo09 (discuss • contribs) 23:15, 3 December 2020 (UTC)

I just found a article on AI taking intelligence tests which I think may be useful for the conclusion conflict. I will summarise my reading before adding it into the main Wikibook tomorrow. Just to inform you guys to keep in mind if you begin any conclusion writing. Camjo09 (discuss • contribs) 18:11, 8 December 2020 (UTC)

General notes
As a question for thought - Do we need a clearer purpose of why we have chosen to identify this? Why is the fact we have identified conflict between different discipline's evidence in measuring intelligence a important issue? Why is it a important interdisciplinary issue I guess?

As discussed in last week's meeting I thought I would add in the rough idea conflict we arrived at: Issues between measuring intelligence by IQ and how that conflicts with cultural differences. Furthermore how that conflicts with the computing notion that humanity isn’t inherently needed to have intelligence, which seems to invalidate needing to accommodate the inherently human variable of culture as evidence in intelligence tests. Camjo09 (discuss • contribs) 23:15, 3 December 2020 (UTC)

--

Yeah you raise good points about being explicit about exactly what the issue is and why it is important. Perhaps we can talk about why it is important as a social issue - quite heavy debates exist around the issue of measuring intelligence, its purpose, its validity, and its consequences. The difficulty would be relating this to whether or not computers can be intelligent, so I don't know if maybe anyone has any ideas on angles we could take on that?

I think part of the conflict lies in, as we said, how we define intelligence. The 'Facts and Fallacies' chapter in the following book (https://www.cambridge.org/core/books/measuring-intelligence/9EFB4DAA867F73021C145B41D3F49015) provides a helpful outline for distinguishing between IQ, g and intelligence which I think helps to resolve such conflicts, to some degree. The debates over whether or not any particular test can be said to actually provide evidence for intelligence or not seems to heavily depend on distinguishing between these.

As I understand it, the issue of culture seems to relate primarily to the idea that understandings of what constitutes intelligence vary extensively between different cultures, and that different environments require and produce different kinds of mental abilities and strengths and thus produce qualitatively different forms of intelligence. "Cultures designate as “intelligent” the cognitive, social, and behavioral attributes that they value as adaptive to the requirements of living in those cultures." (Sternberg & Kaufman, 1998)... I suppose then the issue is whether computational notions of intelligence, and those measured by IQ, would meet criteria for notions of intelligence produced by very different cultures, as opposed to the more technological Western notions of intelligence? E.g. those that place more emphasis on social intelligence, moral wisdom, etc (many Eastern and African philosophies). Xrcaatnp (discuss • contribs) 18:06, 4 December 2020 (UTC)

--

In terms of computers, as discussed earlier I think the CS view on evidence of intelligence slots in nicely with the anthropological perspective as tests tend to be about natural behaviour, rather than reasoning which can be very subjective across cultures. This may offer a interdisciplinary solution to the social issue.

AI intelligence tests are measuring the replica of the same thing that stats/anthropology are measuring. So this is naturally a interdisciplinary issue as they are all conflicting ways of measuring the same thing. This needs to be solved which leads on to the next point that evidence in measuring intelligence is also a issue in CS as we will not know if AI has met its goals of replicating human intelligence until there is a agreed upon benchmark of evidence for human intelligence. So that plus the social issue are two issues we could mention briefly about why this is a important interdisciplinary topic.

I found this article - https://geneticliteracyproject.org/2017/11/16/measuring-intelligence-iq-problems-present/ - which has a interesting real-world case about 1900s America where IQ tests were used to prove eugenics and how today there is still research on the link between IQ and race. Might be something to briefly mention in why its a important social issue. Camjo09 (discuss • contribs) 12:14, 6 December 2020 (UTC)

--

I have attempted to reduce the word count in my own entry and both the intro and conclusion. I think we still have to reduce it by another couple of hundred words Xrcaatnp (discuss • contribs) 13:13, 13 December 2020 (UTC)

--

I cut down as much as I could in my section without compromising the main points as well as making slight adjustments to the intro and conclusions. I also slightly changed some wording in the computer section to reduce word count, if i messed the section up feel free to change it back. Radpebble (discuss • contribs) 15:17, 13 December 2020 (UTC)

Introduction and Conclusions
Do you think, as mentioned briefly in the meeting yesterday, that we should move the final paragraph under 'Measuring Intelligence in Psychology' to the introduction or to the conclusion? I think they are useful overall points. 94.6.204.149 (discuss) 11:54, 8 December 2020 (UTC) Xrcaatnp (discuss • contribs) 20:03, 11 December 2020 (UTC)

--

I altered paragraph slightly to have less overlap but we can move some of it to the introduction. Specifically the bit about intelligence being hard to define. As for what else should be in the intro, maybe we should briefly talk about the reason why its important to test intelligence, what the tensions are between the disciplines. Maybe we reserve the solution for the conclusion. Thoughts? Radpebble (discuss • contribs) 11:50, 9 December 2020 (UTC)

--

I think it makes sense to have an introduction outlining the conventional use of IQ and psychometric based testing to measure intelligence, along with the points you make in your section about concerns around their validity, and the contestability of definitions. The final summary section on the computer based testing seems like a good starting point for a conclusion, so maybe we could move that over to a 'Conclusion' section. If we have those sections as starting points, over the next week we could go along and make edits/additions as we see fit and check with eachother as we do so? Xrcaatnp (discuss • contribs) 15:38, 8 December 2020 (UTC) 94.6.204.149 (discuss) 15:29, 8 December 2020 (UTC)

--

For the introduction I think we need to address concretely why this is definitely a interdisciplinary conflict, rather than multidisciplinary. Any suggestions for that? Also I think briefly mentioning why its a important conflict to address in the intro, two examples being the knock on affect it has socially and also the difficulty it throws up in knowing if AI has met its goals without having a clear set of evidence of intelligence. But yes I agree with moving the computing summary over to conclusion - so focusing on possible ways to integrate the disciplines (where they compliment each other maybe) and then the conflicts between them there?Camjo09 (discuss • contribs) 17:47, 8 December 2020 (UTC)

--

I tried to make a start on an intro but don't hesitate to change or delete anything! Xrcaatnp (discuss • contribs) 09:54, 9 December 2020 (UTC)

--

I added a bit to the intro with a bit more context. I was also wondering if maybe we should add a line like this "This wikibooks chapter will address conventional testing in psychology and look at the interdisciplinary tensions that arise with the social sciences and computing." just to be very to the point but we may be going over the word limit. Radpebble (discuss • contribs) 14:32, 8 December 2020‎ (UTC)

--

I changed a few sentences in the intro to mention more explicitly the word evidence as the seminar leader pointed out that previous groups didn't follow the specific issue clearly enough. I also changed the part about how a interdisciplinary solution would be helpful as I thought it was important make it clear that we're looking at a interdisciplinary conflict/issue first and foremost - as the marking criteria mentions about being very direct to the topic. Basically I tried to make it more explicit that we are dealing with a interdisciplinary issue of conflicting ideas of what is evidence of intelligence in intelligence testing. I hope this makes sense and that my changes are ok?! Camjo09 (discuss • contribs) 14:25, 9 December 2020 (UTC)

--

Looks good to me!

In reference to reducing the word count, I had some potential ideas... I think the following sentences could be removed without compromising the overall intro: 1) The point about intelligence testing being useful for diagnosing learning disabilities. I think we can make the point that intelligence testing can have valuable practical uses without this. Primarily because this is a contested topic within education and one which we don't address enough. It seems like an area best to be avoided if we are only limited to mentioning it briefly in an intro?

2) The half-sentence 'some prioritising flexibility, others complexity'. Without this it could just be a simple straightfoward sentence: 'Different disicplines have conflicting views on what counts as evidence." It's succint and clear at least and cuts out a few words.

I will also go through and try to remove some words from my own section where possible so that we have room to finish off the conclusion! Xrcaatnp (discuss • contribs) 17:27, 9 December 2020 (UTC)

--

Sounds good yes :) I will also cut down my word count as I just added another section to mine to see if it inspired any more thoughts in either of you re conflicts. I will take out the least relevant information to cut it down once we have got the conclusion in place. Camjo09 (discuss • contribs) 01:40, 10 December 2020 (UTC)

--

I have made an attempt to make some more progress on the conclusion but I'm not sure that I have really addressed what we need to on evidence. Will keep thinking over it too. 94.6.204.149 (discuss) 19:56, 11 December 2020 (UTC) Sorry I keep contributing without logging into my account! Xrcaatnp (discuss • contribs) 20:02, 11 December 2020 (UTC)

--

I am not sure that the second paragraph of the conclusion is necessary as it seems to repeat the same kind of thing as the first paragraph just in a different way. Would it make sense to cut that out of the word count? Camjo09 (discuss • contribs) 16:00, 12 December 2020 (UTC)

--

Yeah sure! Feel free to cut out whatever seems unnecessary or repetitive Xrcaatnp (discuss • contribs) 18:11, 12 December 2020 (UTC)

--

I cut out a bit more of the conclusion because it repeated itself a bit (the part about moving towards a more interactive dynamic environment) because we already mention it in the second sentence talking about the perspectives. Radpebble (discuss • contribs) 14:39, 13 December 2020‎ (UTC)