Do dif­fer­ences in work­ing mem­o­ry and exec­u­tive func­tion­ing affect the use of Stan­dard Eng­lish in work­ing-class children’s speech?

Katie Mans­field 0:00
I’m yeah so hi, every­body. I’m Katie Mans­field. I’m a PhD researcher at the Uni­ver­si­ty of Sheffield, cur­rent­ly in my sec­ond year. If you haven’t guessed already, from the front sort of pic­ture on my first slide, my and sort of the text, obvi­ous­ly, my research lies at the inter­face of lin­guis­tics, edu­ca­tion and psy­chol­o­gy. So it’s a, it’s quite a broad mix of all of them. My back­ground is in lin­guis­tics, but psy­chol­o­gy and edu­ca­tion have always been things that I’m inter­est­ed in.

Yeah, so my talk today is all about work­ing mem­o­ry and exec­u­tive func­tion­ing. And how­ev­er, that plays a role in the use of stan­dard Eng­lish in work­ing class chil­dren speech. So yeah, so we’ll begin. Yeah, so just to pro­vide a brief out­line of what I’ve touched upon today. So just a brief overview of my research just to con­tex­tu­alise the game that I cre­at­ed using the goril­la game builder, just a bit about gam­i­fy­ing psy­cho­log­i­cal tasks, then I’ll show you some quick snap­shots and video, a quick video of the game that I built using the new goril­la game builder. And then we’ll just touch upon some pre­lim­i­nary obser­va­tions that I got from doing the task with the children.

So just a back­ground to my study. Where do work­ing mem­o­ry and exec­u­tive func­tion sort of fit into all this then, how does it fit in with this lin­guis­tics stuff? So it all starts real­ly with the nation­al cur­ricu­lum in Eng­land. What the nation­al cur­ricu­lum stip­u­lates is that all pupils should be taught to speak clear­ly and con­vey their ideas con­fi­dent­ly using stan­dard Eng­lish. One of the main crit­i­cisms of that though, is whether it is tech­ni­cal­ly real­is­tic to expect chil­dren to use stan­dard Eng­lish in these real­ly dis­parate con­texts where in for­mal sit­u­a­tions and infor­mal con­text. Researcher has con­sis­tent­ly shown in lin­guis­tics, that infor­mal con­text favour the use of local ver­nac­u­lar and by ver­nac­u­lar, what I mean is sort of local dialect, for instance, so I’m from Leices­ter, so the local dialect to Leices­ter basi­cal­ly, and it assumes that all pupils will be will­ing to change their way of talk­ing for anoth­er and be able to man­age con­trol dif­fer­ent lan­guage styles. So what it what it does­n’t take into account is this kind of cog­ni­tive dimen­sion of this style shift, in this shift in between the local dialect and Stan­dard Eng­lish, it does­n’t take into that con­text. That kind of idea.

And also with social class as well, we touch upon this work­ing class chil­dren, if we take if we com­pared work­ing class chil­dren to mid­dle class chil­dren, work­ing class chil­dren, the major­i­ty of them will be local dialect speak­ers. And obvi­ous­ly, from this require­ment in the nation­al cur­ricu­lum, they’ll be expect­ed to switch the local dialect forms that they habit­u­al­ly use for Stan­dard Eng­lish ones, in order to suc­ceed at school, and man­ag­ing this mis­match between the vari­ety they use at home with their fam­i­ly with their friends, and that expect­ed of them at school, whilst also hav­ing to com­plete the intel­lec­tu­al­ly demand­ing tasks that the teacher set for them in the class­room. What this could poten­tial­ly do is impact on their pro­cess­ing capacities.

If we look at it from a mid­dle class per­spec­tive, mid­dle class chil­dren, many of them enter the class­room speak­ing Stan­dard Eng­lish. So there­fore, they don’t bear the same bur­den. They come from, they come from this back­ground where speak­ing Stan­dard Eng­lish is the norm. So they don’t have that same sort of cog­ni­tive bur­den that the work­ing class chil­dren would. So the ques­tions that I seek to answer is, is there a poten­tial inter­ac­tion between chil­dren’s work­ing mem­o­ry capac­i­ty and exec­u­tive func­tion­ing, and their style shift and abil­i­ties, and our work­ing class chil­dren placed under increased and inequitable cog­ni­tive load in the class­room because of the require­ment that they use a lan­guage style which is less rou­tine for them.

So gam­i­fy­ing psy­cho­log­i­cal tasks that so if we move on to this, so the three psy­cho­log­i­cal sort of abil­i­ties that I will be test­ing, that I’ve test­ed through the tasks and these are work­ing mem­o­ry capac­i­ty, shift­ing abil­i­ty, and inhibito­ry con­trol? Typ­i­cal­ly, these are mea­sured using quite tra­di­tion­al psy­cho­log­i­cal tests, often by pen and paper or through pro­gram­ming soft­ware, which is spe­cial­ly designed for the run­ning of psy­cho­log­i­cal exper­i­ments. Exam­ples of this could be Psy toolk­it, for instance, which is par­tic­u­lar­ly good for these tra­di­tion­al exper­i­ments. The prob­lems of this though, is that it’s effort­ful. It’s frus­trat­ing, and it’s repet­i­tive for the chil­dren. So what they find is that it’s quite bor­ing, that they can’t sort of engage with it. So it leads to par­tic­i­pant dis­en­gage­ment, and there­fore poor data quality.

This is where the Goril­la game builder comes into play. So I came across this through my super­vi­sor, who’s a

4:58
sort of expert, she works in the psy­chol­o­gy depart­ment at Sheffield. And…hi. Yeah. Yep, one sec­ond. One sec­ond. We’re just mov­ing rooms.

Jo Ever­shed 5:20
Katie, don’t, don’t wor­ry about it at all. I know these things are stress­ful. But every­one’s every­one’s here with you. And what you were say­ing was absolute­ly fas­ci­nat­ing. No, you’re very wel­come. If anyone’s-

Katie Mains­field 5:35
let’s get back to it. Yeah, to get back to it. Yeah.

5:44
Absolute­ly, you’re doing absolute­ly great. Every­one is fine and patient and eager to hear what you have to say next, when you’re ready for us.

Katie Man­field 5:51
Okay, so this is where it kind of leads on then to the Goril­la game builder. And I was intro­duced this for my psy­chol­o­gist, super­vi­sor, at Sheffield who have been, she’s got quite lots of expe­ri­ence work­ing with Goril­la. Quite well trained in it. Some­thing I’ve nev­er heard I’ve nev­er heard of real­ly before. So it’s all com­plete­ly new to me. But I was intrigued to kind of look at the pos­si­bil­i­ties that it could offer to my research. So yeah, so just to sort of intro­duce the game builder to those who have not heard of it before. It’s a new state of the art tool, which allows you to go beyond tra­di­tion­al lab based work. The new game builder tool allows researchers to cre­ate inter­ac­tive games like the one I’ll show you in a moment, which turns for­mal and often repet­i­tive psy­chol­o­gy tasks into quite an excit­ing activ­i­ty that chil­dren actu­al­ly real­ly enjoy doing.

There’s no cod­ing or pro­gram­ming knowl­edge need­ed, which was an absolute dream for me, because I hate cod­ing. And it boosts par­tic­i­pant moti­va­tion and engage­ment with­out com­pro­mis­ing data qual­i­ty, which was anoth­er big advan­tage for me, that kind of drew me to want to use the goril­la game builder. So just to show you to give you a brief, sort of snap­shot of the video that I cre­at­ed. Unfor­tu­nate­ly, the sound is not work­ing for some rea­son, I’m not entire­ly sure why. But this is just a snap­shot of the first part of the game that I cre­at­ed. As you can see, there’s loads of dif­fer­ent stuff. The capa­bil­i­ties of the game builder is absolute­ly bril­liant. You can make it as inter­ac­tive as you want it basi­cal­ly there’s, you can’t you can’t see it now. But there’s back­ground music, there’s audio clips, you can upload, there’s you can bring your char­ac­ters to life, basi­cal­ly, to real­ly add that extra sort of engage­ment with the children.

Yeah, so this is all you can have your own back­sto­ry. So this is all about a bur­glar steal­ing pup­pies, and the chil­dren had to retrieve the keys to unlock the pup­pies. So that’s like the back­sto­ry to this game so there’s three tasks. And they had to basi­cal­ly win a key for each task to unlock each of the pup­pies. And the tasks, the keys were all hid­den in this bur­glars kind of like ware­house where we were stor­ing all his stolen stuff. So yeah, as you can see this, you can ani­mate text, you can ani­mate speech bub­bles, you can have as many char­ac­ters as you want, you can have any back­grounds you want. The capa­bil­i­ties of this is lit­er­al­ly end­less. The more that you play around with it, the more excit­ing it gets. So yeah, that’s just basi­cal­ly the sort of start of the game, how that’s how I set that up.

Yeah, so just to take you through the tasks that I had. So obvi­ous­ly, I was talk­ing about mea­sur­ing work­ing mem­o­ry capac­i­ty. To mea­sure work­ing mem­o­ry capac­i­ty, what I did was a run­ning mem­o­ry span, which is a psy­chol- psy­chol­o­gy, sort of test of work­ing mem­o­ry capac­i­ty. Basi­cal­ly, what it involves is a list of num­bers is read out to the par­tic­i­pant and it will end unpre­dictably. So, the child will not know when it will end. And then what will what it will ask the par­tic­i­pant to do is recall as many of the last items in the list as pos­si­ble. But the catch is it has to be an order. This is quite a real­ly, this is actu­al­ly a real­ly good test of work­ing mem­o­ry capac­i­ty. As you can see, the way I sort of inte­grat­ed that into the game is through a safe. So what I did was told the chil­dren that the key was in the safe, they had the audio clip would play on the in the chil­dren’s head­phones. And what they had to do was enter the code into the safe and click on this blue safe but­ton here to try, to turn the safe and see if it would work. Obvi­ous­ly it was set to have so many tri­als. So until the game end­ed, obvi­ous­ly but they enjoyed this kind of instead of it being like a repet­i­tive task, which it would have been had we not had this sort of inter­ac­tive that had this inter­ac­tive nature. They enjoyed keep try­ing The lock keep try­ing to lock to see if it would work and that was what real­ly sort of engaged them in this task.

And then the shift­ing abil­i­ty. So yeah, to do this shift in abil­i­ties basi­cal­ly their par­tic­i­pants abil­i­ty to shift from one cat­e­go­ry to anoth­er cat­e­go­ry. And to do this I employ the colour shape task which is quite notoriously

10:24
used to test this abil­i­ty in the lit­er­a­ture. And it’s had quite suc­cess­ful results, the colour shape task then what it does, it tests shift in abil­i­ty and by that it means how par­tic­i­pants can resolve inter­fer­ence. So, can they active­ly switch from the one one role of cat­e­goris­ing some­thing to anoth­er. So, in this in this case, what we had was a colour task. So Super­man was with the colours, with the shapes sor­ry, and Won­der­woman was with the colours. And they had to cat­e­gorise them by whether it was a cir­cle or square. And when it was Won­der­wom­an’s turn, they had to cat­e­gorise them as whether it was yel­low, or blue. And in here, this is when we could, so these were all sin­gle tri­als. So the three blocks, we had a shape block colour block, here, we had the switch­ing block as the third block of tri­als, and what the chil­dren had to do, if I just play you this just to show you what it looks like, the fix­a­tion point would appear. And wher­ev­er were the shape, whichev­er side it appeared on was which rule the chil­dren had to cat­e­gorise by. So if the cir­cle appeared, for instance, on Super­man’s side, they had to, they they would auto­mat­i­cal­ly know that they had to cat­e­gorise by shape. If the, if a square appeared on Won­der­wom­an’s side, then they would have to cat­e­gorise by colour.

What I did with this data is, I’m cur­rent­ly in the process of col­lect­ing all of the data. And we’ll be cal­cu­lat­ing reac­tion time and per­cent­age of cor­rect respons­es. Which brings me on to anoth­er point with the Goril­la Game builder. What’s bril­liant about it is that it does it all auto­mat­i­cal­ly for you, you’ll get a spread­sheet with all the data on and it shows you the reac­tion times from par­tic­i­pants. And the accu­ra­cy, you can set up scor­ers, all sorts of things behind the scenes that enables this data to be col­lect­ed, like auto­mat­i­cal­ly and it’s bril­liant. Obvi­ous­ly, it avoids the pos­si­bil­i­ty of man­u­al error as well, which is bril­liant. Anoth­er advan­tage of it.

And then the last test that I did was inhibito­ry, it was test­ing inhibito­ry con­trol. And one of the most used tasks for this in psy­chol­o­gy is the go no go task. What it, what I did was employed two blocks of 50 tri­als. And essen­tial­ly what the go no go test does it for peo­ple that don’t know what it is, essen­tial­ly, it’s a task where you, the child builds up this this response. So for in this case, we had blue dia­monds and pur­ple dia­monds, the chil­dren had to press the space­bar every time they saw the blue dia­mond. And then because they build up this auto­mat­ic response, where they were expect­ing a blue dia­mond, when a pur­ple dia­mond appeared on the screen, they had to con­trol them­selves and stop them­selves from press­ing the space­bar for the pur­ple. So yeah, that’s how it kind of tests inhibito­ry con­trol. Yeah, so two blocks of 50 tri­als. And again, I’m cur­rent­ly work­ing through col­lect­ing the rest of the data for this. And we’ll be look­ing at reac­tion time and accu­ra­cy. Just to show you a quick video of how this one works. So again, we have a fix­a­tion point if it wants to play. If it’s going to work, yeah, so we have a fix­a­tion point and then the dia­mond will appear. And the chil­dren have to react accordingly.

Just some obser­va­tions then to fin­ish off with because obvi­ous­ly, I’ve not got any sort of data to present unfor­tu­nate­ly today. In terms of things that I’ve noticed, whilst car­ry­ing out the game with the chil­dren, the first one no reduc­tion in data qual­i­ty. I think that’s one of the biggest things for researchers like your­selves who are prob­a­bly inter­est­ed in poten­tial­ly using a game builder like this. The con­cern I def­i­nite­ly had at the begin­ning was whether there would be a reduc­tion in the data qual­i­ty from, obvi­ous­ly gam­i­fy­ing some­thing does it would it alter the qual­i­ty of the data I would be able to col­lect, and it hasn’t.

And the sec­ond biggest thing for me was par­tic­i­pant engage­ment. So par­tic­u­lar­ly what I noticed was obvi­ous­ly in a in an aver­age class, you always get some chil­dren who’s poten­tial­ly, who pos­si­bly don’t real­ly have the best atten­tion spans, they prob­a­bly have low­er atten­tion spans than oth­er chil­dren. And what I noticed was the chil­dren that I was par­tic­u­lar­ly con­cerned about doing this task, the task with, whether they’d be able to sit in one sit­ting and com­plete the game, were the ones who actu­al­ly were the ones that I should­n’t have wor­ried about at all. They sat down, they were com­plete­ly enthralled by the game. They thought it was great. They just want­ed to get, which brings me on to my next point. They just want­ed to com­plete the game again and again and again in their free time. I’ve had the chil­dren com­ing up to me whilst I’ve been work­ing at the school for the last cou­ple of weeks, since we’ve been doing the game. And they’ve been say­ing, Oh, can we do the game today, can we play the game again I want to beat my score. So that just kind of tells you all you need to know real­ly about this par­tic­i­pant engage­ment and how gam­i­fy­ing, some­thing that would be tra­di­tion­al­ly quite bor­ing, quite repet­i­tive, actu­al­ly can be made into some­thing that a child will enjoy.

15:25
And just the next step, real­ly, in terms of how I’ll be link­ing this data to the rest of my project that I’m doing, I’ll be doing a speech based exper­i­ment. This is more of the lin­guis­tic based stuff. And what it’s designed to do is induce vary­ing lev­els of cog­ni­tive load. They basi­cal­ly has a read­ing and com­pre­hen­sion task that set as the con­trol task, and then it’s got four dif­fer­ent con­di­tions, as you can see, so dif­fer­ent demands are placed on the chil­dren. So you’ve got the whether that we’re told to speak in their home lan­guage, whether they’ll be told to speak in the school lan­guage. And in these two con­di­tions, whether there’ll be in these two con­di­tions, sor­ry, two and four, whether there’ll be an addi­tion­al task they have to com­plete, which will be an audi­to­ry based task, as well as the com­pre­hen­sion task. And what I’ll be doing is try­ing to draw some cor­re­la­tions between the data that’s col­lect­ed from the Goril­la game that I’ve cre­at­ed, and this speech based exper­i­ment data to find out whether the chil­dren’s work­ing mem­o­ry and exec­u­tive func­tion scores are relat­ed to their per­for­mance in the speech based exper­i­ment, and start to explore those poten­tial impli­ca­tions for edu­ca­tion­al pol­i­cy and practice.

That’s me fin­ished. Thank you for lis­ten­ing. And thank you for putting up with my minor pause while I had to change room. I’ve includ­ed my Twit­ter han­dle here, just for any­body who’s inter­est­ed in in my research or inter­est­ed in how I went about build­ing the game, and could poten­tial­ly be inter­est­ed in doing that them­selves. So yeah, you can catch any updates on my PhD on Twit­ter. Hope­ful­ly, there’ll be some, some high­lights and stuff with my data soon. So thanks for lis­ten­ing. And I’ll be hap­py to take any questions.

Jo Ever­shed 17:08
Katie, that was phe­nom­e­nal. While peo­ple are putting their thoughts togeth­er into ques­tions in the q&a, I just want­ed to say I get if you were my PhD stu­dent, I’d be tremen­dous­ly proud. That’s a phe­nom­e­nal amount of work you’ve done. And it’s extra­or­di­nar­i­ly dif­fi­cult doing work with kids in schools, psy­cho­log­i­cal and you’re man­ag­ing to clear­ly jug­gle a very diverse set of work, and you’re doing it with tremen­dous grace. So well done. If any­body does have any ques­tions for Katie, please put them into the q&a. That would be real­ly inter­est­ing, you’re get­ting lots of pos­i­tive feed­back here. Awh, is that the super­vi­sor, super­vi­sor, lots of peo­ple say­ing it’s very inter­est­ing work. And what a great approach. It’s one of the things we’ve heard a lot from researchers like you as well, the capac­i­ty to go and test kids with spe­cial needs. With a gam­i­fied tool. It’s just com­plete­ly trans­for­ma­tive. And there are so many par­ents and car­ers and teach­ers out there who actu­al­ly have very weak under­stand­ing of what their chil­dren with spe­cial needs are able to do where they are at the moment. And if you could get decent mea­sures, and then be able to track them over time, because you could send that game off to them, then then that would be a tremen­dous ben­e­fit. Yeah. There’s a ques­tion here from Kim­ber­ly Beau­mont. Amaz­ing talk, Katie. Just curi­ous, how were the chil­dren respond­ing in the switch task? dif­fer­ent but­tons for the cir­cle square ver­sus the red blue? Like, what were the con­trols here?

Katie Mans­field 18:38
Yeah, so for the colour shape task. So in the game builder, what they, it’s real­ly clever how you can set it up. So you can set it up, set up a key­board response. So the way this par­tic­u­lar task was set up is if the chil­dren were respond­ing to it in the they obvi­ous­ly had the sin­gle tri­als of the colour tri­al and the shape tri­al. So in the shape tri­al, if they, if a cir­cle appeared on the screen, they would press M screen, they would press the J key, then what would hap­pen in the next tri­al for the colours, they would press F if the shape was yel­low, and it com­plete­ly ignored the shape just look­ing at the colour this time? And they would press J if it was blue. And then obvi­ous­ly, we’ve got this thing because they’ve learned two dif­fer­ent rules. Can they switch between those rules in the in the mixed, obvi­ous­ly, the switch­ing part, the switch­ing block of the tri­als? Where, depend­ing on which side it appears, can they auto­mat­i­cal­ly retrieve that role that they’ve learned and know whether it’s going to be the obvi­ous­ly the J key or the F key?

It was quite a tricky one, to be hon­est. It was quite a tricky one for them to get going with it. Once they had the prac­tice tri­als though, because I did include some prac­tice tri­als. I should have men­tioned that just to get them famil­iarised with the sort of the con­cept to have what they were meant to be doing. They flew through it, they were absolute­ly fine. So it was it was nice to see how much they enjoyed just play­ing around with it. Because I was real­ly wor­ried to start because this is some­thing quite new to me I’ve not, as my super­vi­sor will prob­a­bly tell you, I’m not in any way famil­iar with Goril­la, it’s been some­thing com­plete­ly new from scratch to me, that I’ve had to do. And going back to Decem­ber, when I first start­ed, I was hav­ing an absolute melt­down think­ing, I’m not gonna be able to do this, there’s so many dif­fer­ent but­tons, there’s so many dif­fer­ent capa­bil­i­ties that this this piece of soft­ware offers. I don’t even know where to start. I took some time away. And I’m so glad I said, say to every­one at the uni­ver­si­ty now, I’m so glad that I did­n’t just walk away from the game builder and actu­al­ly got the licence and sat and per­se­vered a bit because it was one of the best things that I’ve done in my PhD. It’s real­ly helped me to take this dif­fer­ent approach to take this dif­fer­ent per­spec­tive and actu­al­ly pro­duce some thing of this qual­i­ty, I nev­er thought that I’d be able to pro­duce some­thing like this, that col­lects the data in such a, just an effi­cient way, like the way the data is com­ing out in the spread­sheet at the moment. It’s bril­liant. It’s so easy. And as I said in my talk, it just avoids these prob­lems of man­u­al errors. That’s so many of us when we’ve got so much work to do. It’s so easy to just make these sil­ly errors that you don’t even realise you’ve done and but the, goril­la it just sim­pli­fies so much stuff for you. It’s bril­liant. Yeah, it’s absolute­ly brilliant.

Jo Ever­shed 21:30
well that’s fan­tas­tic. We, and I’m so glad you per­se­vered as well. I think what you’ve man­aged to cre­ate here is tru­ly impressive.