On prosociality and language, with Dr Umar Toseeb

Dr Umar Toseeb is a Lecturer in Psychology in Education at the University of York. He is interested in children's mental health and wellbeing, specifically children with developmental differences such as developmental language disorder and autism spectrum conditions. During this podcast he chats to Sue about his recent paper on how prosocial behaviours (helping, sharing, caring etc.) develop during childhood and how they are related to mental health and wellbeing in children at risk of developmental language disorder.

You can follow Umar on Twitter here.

The paper discussed in this episode is:
Toseeb, U., & St Clair, M. C. (2020). Trajectories of prosociality from early to middle childhood in children at risk of Developmental Language Disorder. Journal of communication disorders, 105984.

Transcript

Sue: 0:09

[Podcast jingle][ringtone] Hello? Oh, it is recording. I see the little figure. Okay, great. I will do my little spiel and then I'll introduce you. Nice. Okay. Here I go. Hi, I'm Sue from the Salvesen Mindroom Research Centre at the University of Edinburgh, and we are recording our 10th PsychologiCALL! Very exciting! Um, today I'm talking to Dr. Umar Toseeb from the University of York, and he's going to be talking to me about a paper that was looking at trajectories of prosociality from early to middle childhood and children at risk of developmental language disorder. Um, so welcome, Umar! It's really nice to talk to you today.

Umar: 0:51

Thank you Sue, thank you for having me!

Sue: 0:53

Not at all. So, um, I'm sure there's some words in that title that we will want to unpick in a moment, but perhaps you could start by just telling me, um, what your kind of key discovery was. Do you think in this particular piece of work?

Umar: 1:10

Well what we found was that children at risk of this disorder are[inaudible] quite prosocial. So they have been sharing and caring. And it's not something that they necessarily struggle with. They do do slightly worse than children without this disability, but on the whole, they're still within the sort of what we call the"normal range".

Sue: 1:29

Great, so we've got a good news story here about, um, prosociality. So you mentioned that some, um, examples, but could you expand a little bit on what we mean by"prosocial" when we're talking about that in psychology?

Umar: 1:46

Yeah, so, um, it's quite a broad term, but within our study we were specifically interested in, um, when children share toys, or how[inaudible] maybe upset or hurt, um, or sharing- or showing caring behaviours towards other people.[inaudible] measures, so we asked parents to indicate which prosocial behaviours their child showed through a given timepoint.

Sue: 2:12

Amazing. Um, and so then you mentioned as well these were kids at risk of developmental language disorder, and that's, I know quite a common condition, but relatively, um, unknown, I think, in the public domain. It's certainly not talked about as much as things like autism and ADHD. So perhaps you could just take a moment to describe what developmental language disorder is and what you mean by children at risk in this context.

Umar: 2:46

Yes, so a developmental language disorder is uh...a disorder when children have under-, um, difficulties understanding and using spoken language. In effect, approximately 5 to 7% of shool-age children, so it's much more prevalent than autism. But a lot of the time it goes unrecognized, just because, um, the children might exhibit other types of problems which are not necessarily seen as language, so behavioural problems or other the types of mental health difficulties. And the reason why we've called it"children at risk of developmental language disorder" is because we used a... the sample we used we weren't able to do a clinical diagnosis of developmental language disorder, we used an existing sample[inaudible] later on. And we couldn't be sure if the children had developmental language disorder...

Sue: 3:33

Right.

Umar: 3:33

But with the measures that we had we could say that they were at risk of developmental language disorder.

Sue: 3:37

Right. Okay. And so, so tell us a little bit then about what motivated you to be looking at prosociality in this group. Is this something that has been much studied before, or is it quite new territory?

Umar: 3:52

Well it's relatively new. We know that children with developmental language disorder tend to have poor mental health compared to without developmental language disorder, and we also know that prosocial behaviours are associated with better mental health in children in general. So what we were interested in was to see whether prosocial behaviours in children with developmental language disorder, how they develop, um, and whether that development was associated with fewer mental health difficulties. And that's important because that is the first step towards understanding whether interventions focused on prosocial behaviours can be used to reduce the prevalence of mental health difficulties within children with developmental language disorder.

Sue: 4:30

Right. Right. So, so it's all about preventing mental health difficulties and, and sort of investigating whether prosociality could be a good target in that kind of process.

Umar: 4:42

Yeah, absolutely. So I suppose, the good news about language difficulties is that there are[Inaudible] a number of things that can help, and we just pick up on prosocial behaviours, because there's some evidence that in the general population prosocial behaviours are helpful, and we were interested to see whether they are particularly helpful for children at risk of developmental language disorder. And I suppose, it's more interesting for children with developmental language disorder because, because of their problems with using language, they might not be able to access other types of mental health supports that are reliant on the use of language. So if we can try and find things, try and identify things that don't necessarily rely on language that could also help support mental health, then it might be particularly helpful for children with DLD.

Sue: 5:27

Right. Yes. That makes a lot of sense. And so, um, so tell me a bit about the methods that you use to look about. You've mentioned trajectories in your title, and I'm wondering if you managed to pull off a longitudinal study, but that's, that's obviously very hard to get funding and resources for. So tell me about what you did.

Umar: 5:46

So we were quite lucky. So we used data from an existing cohort, so we've got funding from the, uh, UK-based Millennium Cohort study. Um, and what that is is the... Around the year 2000- 2002, um, the Millennium Cohort group was created as a representative sample of about 19,000 children who were born in the UK. And then they followed those children throughout childhood, and there were multiple times when parents were asked to complete questionnaires and children were asked to take part in certain tasks. Um, and what we did was we took the data from the age of 5, so when the kids were five years old, and we completed a brief language assessment, and their parents were also asked to report on any language concern that they had about their children's language development. And then we used that data from those language questions, and we identified the children at risk of DLD. And then we also use the longitudinal data on prosocial behaviour from the age of 5, 7 and 11. So they were the same families that were followed over a period of time. So we did have a longitudinal approach[inaudible] over a number of years.

Sue: 6:57

Fantastic. Yeah, yeah. Sorry.

Umar: 6:59

Yeah, I was going to say, and then once we identified the group of children who are at risk of DLD, um, then we looked to see how their trajectory of prosocial behaviour compared to children without DLD, or sort of how groups of children change differently over time irrespective of whether they have DLD or not.

Sue: 7:18

Alright, so can you tell me a bit more about some of the patterns then that emerged from the data?

Umar: 7:26

Um, yes. Um... So we used a technique called, um, latent class growth curve analysis, which takes the group or population for example, and identifies how some groups of children change in a, in a similar way over time. Um, and what we found was... There were 4 groups of children, and the core group was"stable high", so what that meant was children who started off quite prosocial at the age of 5 and continued to be prosocial at the age of 11. There was a decreasing group, so children who were quite prosocial at the age of 5, but by the time they got to the age of 11 they weren't so prosocial. Um, and then we had an increasing group, children who were not very prosocial at the age of 5, and by the time they got to the age of 11 they were highly prosocial. And then we got a group of children consistently quite low, um, relative to the others they were at the lower end of the prosocial scale. All of the groups within again what we call the"normal range". Um... There wasn't a group that was at the point where it was a call for concern, um... They're just varied[inaudible] down[inaudible] within the normal range.

Sue: 8:40

And so, so I've never done one of those growth curve analysis, but if I, if I understood it correctly, what you're doing then is comparing the shape of the pathway that the children are following rather than where they are at five years old and where they are at seven years old and where they are at 11 years old. Right? It's, it's all about the sort of, the trajectory, as you say, in your paper, you know, the kind of journey that they're on, is that right?

Umar: 9:07

Yeah, absolutely. So, it's a data driven approach, so, um... the... The model takes the data and looks to see how the groups of children are changing over time. It allows the children to increase over time, decrease over time, or stay the same[inaudible]. It's not necessarily looking at age 5, age 7, age 11. What is the growth or the change[inaudible].

Sue: 9:34

And so what do you think of the implications for this work? You know, what do you think we can learn from this, I suppose specifically in relation to developmental language disorder, but perhaps also more generally about, you know, prosociality and the development of that in children?

Umar: 9:52

Well, I think... For me what's important is, we learned that prosocial behaviours are potentially quite important for children with DLD. So[inaudible] what we found was that children with DLD who are prosocial tend to have fewer mental health difficulties. And if future research can establish that there is a causal effect between prosocial behaviour and mental health difficulties, then interventions aimed at promoting prosocial behaviour might go on to reduce mental health difficulties. And like I said earlier it's... It's particularly important for children with DLD who might not be able to access support for mental health that relies on language. So identifying prosocial behaviours as a potential point of intervention,[inaudible], I think that's the key things I would take from this study.

Sue: 10:43

And what about that group of children who have got sort of declining prosociality over time? You know, do you have any, any, um, insights or thoughts about what might be causing that to happen? I could imagine that there could be maturational effects where you know, that the, the... The level of behaviour that is required to be thought of as prosocial when you're 5, you know, the standards rise around you. So those children are not necessarily becoming less prosocial, they're simply kind of staying in the same place while everyone else's expectations move on. You know, do you have any thoughts about who that group is and, and what's there, what's the story behind that?

Umar: 11:30

Um, well, I should start by saying that was a very small proportion of our overall sample, so that was 6% of the sample, and we followed that decreasing to slightly lower trajectory. Um, and for the most part, everybody else was in the rest of the trajectory. Um, and I think, although we didn't look specifically at what that group consists of, um, what we did find was important[inaudible], the levels of prosocial behaviour that changed over time. Um... It was your current, it's the current- it's the current levels of prosocial behaviour that are the most important for the[inaudible] rather than your predicted growth[inaudible]. And to have a decrease in levels of prosocial behaviour, in earlier childhood their levels of mental health issues might be more[inaudible] of their prosocial behaviour[inaudible] levels of mental health issues[inaudible].

Sue: 12:33

Hmm. Interesting. Um, so, um, so what do you want to do next? Have you got another, um, Millennium Cohort analysis up your sleeve, or are you moving this forward in a different direction?

Umar: 12:48

So, um, I think, the next step will really be about establishing causality, so we know how these things are related. And we also now know that these two things are related to mental health and prosocial behaviour are related in samples of children with DLD, and we've looked at[inaudible] in a young age group. The next step is to establish whether increases in prosocial behaviour leads to decreases in mental health difficulties. And[inaudible], ideally, we want to do some sort of randomized controlled trials, but I feel like we're quite a bit away from that. And I think we would need more research to really try and understand, um, what-, what's the extent of the association between the two areas[inaudible], whether it varies[inaudible] for different languages within DLD, within different subtype. Um... Then I suppose the next step really will be to carry on using the existing data sets that we have and, and use some causal inference modelling, so some more[inaudible] techniques that allow us to draw more causal inference before we can do some randomized controlled trials.

Sue: 13:57

Um, and I think, I think that reveals a really important point, actually that's maybe worth emphasizing, which is, you know, people can be frustrated at the slow pace of translation of research into practice, right. But at the same time, you know, a good, well powered randomized controlled trial costs a huge amount of money. It requires investment from a large number of people. And especially of course, all the participating families and young people and so on and to, to go into that prematurely before having fully, you know, sort of, um, really ascertain the right target and the right population is really irresponsible. So I think you're, I think you're right to be taking that sort of step by step approach. But, but I do, I do sometimes feel like this is one of the things that, that can be frustrating for people outside research, just to sort of see how, you know, incremental[laughs], um, our progress is.

Umar: 14:58

Yeah. And I think also what's important there is prosocial behaviour is just one aspect that could potentially be targeted, and I suppose what we really need to figure out is to what... How much of a difference would it make to, um, mental health difficulties in this sample. Because it might be that there are other things which are cheaper to um, or more efficient or quicker, or easier to implement that would have more of an effect. Um... So of course, resting the research on prosocial behaviour and mental health[inaudible], but also trying to investigate the other things that could help as well. Um, so we could have a fully formed, fully informed intervention, rather than one[inaudible] one of the aspects that we've been looking at.

Sue: 15:46

Um, absolutely. So, um, Umar that was really fascinating. Before we finish, I always ask people if they've got any advice for any students or early career researchers who might be listening and perhaps, uh, uh, given the, the global situation, perhaps be feeling particularly in need of some words of wisdom at the moment. So have you got anything that you would like to share with those people in our, in our audience?

Umar: 16:16

Yes, so I would say that if you have a knack for stats, um, then try and learn some new techniques and develop the skills in this area, because personally, I just think it's... statistics are just a lot of fun. Um, and secondly, I think if you're able to do analysis over and above, like the standard ANOVA or regression, for example, then it probably-, I feel like it opens a lot more doors for you, as well as employment opportunities, collaboration, but just being able to get on and use existing data and do some analyses and write a paper, which I feel like helps your career anyway. Um, so yeah, I would say if you're into it and you like it, learn some statistical modeling techniques. And it doesn't have to be anything extravagant or exciting, it's learning something new that you don't already know always helps.

Sue: 17:06

Yeah. I completely agree. I think that's an excellent piece of advice. And I think as well, it's, it's not just about, um, your own analyses and your papers, but also your ability to make sense of the literature. Because as those analysis-, analyses become more and more a part of how we're kind of exploring, um, understanding development, you know, you really need the critical skills to understand them and to, you know, to understand the assumptions that underpin them and the, you know, how, how robust the, the results are, and any kind of limitations on them. And that's something where I feel sometimes at a disadvantage because I don't have super sophisticated stats knowledge, and I don't now have a huge amount of time for developing that knowledge. So, yeah, I think that's a great piece of advice.

Umar: 17:57

[Inaudible] to finish on one final thing, one of my previous mentor once said to me"just try and learn a new, a new analysis techniques in each paper that you write", which I think is really optimistic, and I just don't do that and I can't do that. But I try and try and do it for like any paper that I think we could try and apply a new analysis technique to. And in that way it gets you to learn in as well and then applying it there, rather than just learning and not do anything with it.

Sue: 18:20

Yeah, that is a great piece of advice. As you say it's a bit of a stretch, but you know, it's good to have goals.

Umar: 18:28

[Laughs]

Sue: 18:28

So thank you so much for your time Umar, this was fantastic. Um, for anyone listening, you'll be able to find out more about the work we've been discussing by following the links on the podcast page, which is at ed.ac.uk/salvesen-research. And it just remains for me to thank you Umar for joining me today and thank everyone who is listening. Thank you bye![ringtone] Okay we did it! I thought that went quite smoothly![Podcast jingle]