Does the speed of autom­a­ti­za­tion pre­dict dif­fer­ences in children’s lin­guis­tic ability?

Ash­ley:
Hi every­one and wel­come to my talk. So my name is Ash­ley. I’m a PhD stu­dent at the Uni­ver­si­ty of Birm­ing­ham, and I’m part of a research group led by Pro­fes­sor Ewa Dabrows­ka. As a group, we research lan­guage acqui­si­tion and enter­tain­ment. Today, I’m just going to give you a lit­tle back­ground about my research, but I’m going to talk more about how I moved my research online rather than my research itself. So my research inves­ti­gates whether there is a rela­tion­ship between the speed in which chil­dren autom­a­tize a com­plex cog­ni­tive skill, and we use the Tow­er of Hanoi puz­zle and lan­guage learn­ing. And we’re com­bin­ing two the­o­ries. We’re com­bin­ing the skill acqui­si­tion the­o­ry and the usage-based mod­el of lan­guage acqui­si­tion. And in my research, I’m inves­ti­gat­ing two groups, typ­i­cal­ly devel­op­ing chil­dren and chil­dren with devel­op­men­tal lan­guage disorder.

Ash­ley:
So why did I move my research online? I think the answer to this is obvi­ous, and the chal­lenge I faced was obvi­ous­ly the pan­dem­ic. I received my eth­i­cal cur­rents in March last year and was real­ly excit­ed to get going. My research design was me going into pri­ma­ry schools and doing my research with chil­dren rough­ly sev­en to nine years but had to go back to the draw­ing board and think what to do. We came up with the solu­tion of mov­ing our research online even­tu­al­ly. I think for some time, we were unsure how long it would be before schools reopen, but these answers were not forth­com­ing. I did actu­al­ly get into schools and then had to stop, so it was a bit of stop-start for awhile. But I was design­ing a task on anoth­er plat­form. And I hap­pened to go to the BeOnline Con­fer­ence last year. And there, I was intro­duced to Goril­la and decid­ed to move the design of this task onto Goril­la and had some amaz­ing sup­port from the team of Goril­la and spe­cial shout-out to Will Webster.

Ash­ley:
I was also very for­tu­nate to win British Psy­cho­log­i­cal Soci­ety Post­grad­u­ate Rapid Research grant, and this helped me with some scripts in con­sul­tan­cy. I did­n’t script myself. That’s some­thing I think I real­ly need to learn, but you don’t have to script to use Goril­la, and that is the joy of it. And so we decid­ed to move my research online but for myself as the researcher to be present, so very much in line with the talks that you’ve just heard, I guid­ed the chil­dren through each of the tasks. I resub­mit­ted my changes for ethics approval, and then I recruit­ed my par­tic­i­pants. So all in all, I recruit­ed 101 par­tic­i­pants, and I did my research via Zoom over three sessions.

Ash­ley:
I men­tioned to you that we use the Tow­er of Hanoi puz­zle to mea­sure cog­ni­tive skill learn­ing, and my super­vi­sor and I decid­ed very much that we want­ed to use a phys­i­cal puz­zle, so this is the phys­i­cal puz­zle, for chil­dren. So we decid­ed to post every child that took part a puz­zle, and this is exact­ly what the puz­zle looked like. They received it in this love­ly box, wrapped in a rib­bon. And we said to par­ents, “Please, please, please don’t open the puz­zle before the first ses­sion.” So I would meet the child on Zoom, and they would unwrap a lit­tle present, which was real­ly fun for them. And then I would give them instruc­tions as to how to do the task.

Ash­ley:
The task itself, I’ll just show you this while I talk, the Tow­er of Hanoi, if you’re not famil­iar with it, we used four disks of the puz­zle and the job for the child is to move the four disks from the left-hand side to the right-hand side, but they can only move one disk at a time and they can’t put a larg­er disk on top of a small­er disk. And it’s ide­al for mea­sur­ing cog­ni­tive skill learn­ing because you mea­sure the dif­fer­ent stages as the child pro­gress­es through the puz­zle. And it’s a tan­gi­ble puz­zle. It’s real­ly fun for kids to keep them engaged and keep them moti­vat­ed. And it’s got a recur­sive struc­ture, so they’ve got to keep sub-goals in mind. And we also used a sec­ondary task to mea­sure any inter­fer­ence effects and to see whether chil­dren had real­ly pro­ce­du­ral­ized the solu­tion to the puzzle.

Ash­ley:
The next two ses­sions, you’ll prob­a­bly be very famil­iar with these, so I won’t read them out. And we used a nar­ra­tive task and expos­i­to­ry dis­course task and var­i­ous lan­guage mea­sures. So, look­ing at gram­mar, vocab­u­lary, recall­ing sen­tences. And in cog­ni­tive tasks, we use ravens. Two tasks you may not be famil­iar with is the back­wards col­or span task devel­oped by Nick Rich­es, which is like back­wards dig­it span, but using balls and a tube. And the task I designed on Goril­la, which is embed­ded triplets. It’s a sta­tis­ti­cal learn­ing task designed on an orig­i­nal design by Arci­uli and Simp­son. So these were two sep­a­rate ses­sions that I ran again on Zoom.

Ash­ley:
So my tips and thoughts about mov­ing your study online, ethics is obvi­ous, and you’ve also got to think about the task demands when you move your tasks online, espe­cial­ly with work­ing with chil­dren. And this is the length of the tasks and how dif­fi­cult they are. Does the researcher need to be present? So in my case, I was. And I could guide the chil­dren through each of the tasks, but if you’re not going to be present, then real­ly clear instruc­tions. Again, this is being touched on, but for me, par­ents and car­ers as co-researchers was absolute­ly invalu­able. Par­ents were real­ly on board with the research and real­ly inter­est­ed. So they helped me do things that I could­n’t con­trol over­see­ing, like the envi­ron­ment, keep­ing the envi­ron­ment qui­et, keep­ing chil­dren engaged online. There’s lots of things that you can do to keep chil­dren engaged. I did send stick­ers and reward charts.

Ash­ley:
My col­league, Mag­da, designed avatars with chil­dren just at the begin­ning of the study. And again, it’s the set rewards like stick­er charts, and we obvi­ous­ly com­pen­sat­ed chil­dren for their time. And as peo­ple before me have said, pilot your study exten­sive­ly, so I pilot­ed my study in per­son. And then again, online, when I moved it onto an online format.

Ash­ley:
Things to bear in mind, do you need any spe­cial­ist tech­nol­o­gy? I was advised to use a dig­i­tal visu­al­iz­er and it was absolute­ly invalu­able for my research. And to help me share things online through screen share. Do you need par­tic­i­pants to wear head­phones? Also think about their access to tech­nol­o­gy. Some fam­i­lies might only have a mobile phone in the house. So do you need to spec­i­fy whether your tasks are done on a com­put­er? Can they also run on a tablet?

Ash­ley:
Wi-Fi sta­bil­i­ty. We’ve obvi­ous­ly not got much con­trol of, but I did have this occa­sion­al­ly where some­times we would have to stop a call and restart, but it real­ly was not a major issue for me. I recruit­ed online. So I recruit­ed through Face­book, through spe­cial­ist groups, through Twit­ter, through word of mouth, and through groups like Chil­dren Help­ing Sci­ence. And it was very suc­cess­ful for me. There is obvi­ous­ly sys­tem­at­ic bias. You reach in a much wider group of par­tic­i­pants but per­haps you can’t con­trol who decides to take you up on your study.

Ash­ley:
I’ve men­tioned that the oth­er joy is you’re not con­strained to the school day. So I lit­er­al­ly did my research for sev­en days a week. When­ev­er par­tic­i­pants could meet me is when I would meet chil­dren online. Sched­ul­ing mul­ti-ses­sions. I did­n’t actu­al­ly have any dropouts. I was very for­tu­nate. All chil­dren came to all three ses­sions and I also used an online sched­ul­ing sys­tem, which would send par­tic­i­pants a Zoom link and a reminder for each of their ses­sions, and I found that real­ly helpful.

Ash­ley:
So just in con­clu­sion, we are con­tin­u­ing with our online devel­op­ment. We’ve now moved the Tow­er of Hanoi puz­zle onto Goril­la. Thanks again to Will Web­ster who’s helped us pro­gram this onto to Goril­la. And at the moment, we’re using it with teenagers and with adults and not yet with chil­dren. But as a gen­er­al clos­ing, all of our stud­ies in our lab are now run­ning on Goril­la. It’s helped us to con­tin­ue our research dur­ing the pan­dem­ic, and we can reach a wide range of par­tic­i­pants. It’s amaz­ing. It helps us save time. We col­lab­o­rate with each oth­er. I moved the oth­er researchers at oth­er uni­ver­si­ties, and it’s real­ly helped us all gain a wide range of new skills. So thank you very much for lis­ten­ing to me, and I’m look­ing for­ward to the rest of the con­fer­ence. Thank you.

Speak­er 2:
Thank you very much, Ash­ley. There’s a quick ques­tion about socioe­co­nom­ic sta­tus. How did you match patient groups for that? So this is from Kirsten Abbott Smith, who has found that it’s less of a prob­lem when you’re test­ing in schools because there’s often more vari­ety in the schools. But if you have a vol­un­teer basis, mid­dle-class par­ents tend to be more like­ly to vol­un­teer. How have you kind of nav­i­gat­ed that? Or you’ve kind of hint­ed at this already?

Ash­ley:
Yeah, I did hint at it. I think, to be hon­est, this is where I hint­ed at the sys­tem­at­ic bias. I think a lot of par­ents who will take part in your study are per­haps par­ents who are quite moti­vat­ed for their child to do this, and they help the child to be moti­vat­ed. And I do agree it’s a dif­fi­cult thing to con­trol for, and less of an issue if you’re doing in-per­son research in schools. It is a slight wor­ry. You try as hard as you can to reach as many peo­ple if you pos­si­bly can, but it’s not always possible.

Speak­er 2:
And going for­ward, hope­ful­ly we’re not always going to be in a sit­u­a­tion where it’s not pos­si­ble to be doing things in school. So maybe a blend­ed approach could be real­ly valu­able in the future. Thank you very, very much, Ash­ley. And remem­ber, you can keep ask­ing ques­tions in the Q&A and actu­al­ly the oth­er speak­ers will pick up on that.