PSYU2236 BIOPSYCHOLOGY AND LEARNING 1Practical One: Research ReportCan we learn from watching robots?PSYU2236 BIOPSYCHOLOGY AND LEARNING2Today’s practical will involve:– Providing a background to physical and observationallearning.– Outlining some theories of how or why observationallearning works.– Describing data from a recent experiment thatcompared how well people learn from watchingdifferent models – human versus robot.– Explaining where you can find the data, as these datawill be used for your research report.Practical 1 Overview3For a background to motor learning and observationallearning of motor skills, see Chapter 11 in Mazur.We can learn motor skills by physically practicing and wecan learn by watching other people. But to what extentcan we learn by watching robots?Practical 1 Overview4BackgroundWe can learn new skills by physical practice and by watchingothers perform (observational learning)5BackgroundIn certain contexts, observational learning is highlyadvantageous over physical practice6The ‘like-me’ hypothesisMeltzoff, 2007“The ‘like-me’ nature of others is the starting point for socialcognition”7The mirror neuron systemGallese et al., 1996; Fogassi et al., 2005Human fMRIGrèzes & Decety, 2001; Gazzola & Keysers, 2009Although these findings are noteworthy,this was not a new idea:James (1890)Jeannerod (1994)Prinz (1990; 1997)IFGIPLMonkey recordings8Mirroring and learningBlandin et al., 1999; Gog et al., 20099What about the model’s identity?Does it matter who we are learning from?Experts vs NovicesA model thatdemonstrates masteryvs a model that developsskill over timeBoth types of modelare effectiveAnd maybe acombination is best(e.g., Rohbanfard &Proteau, 2011)10Does it matter who we are learning from?What about learning from non-humans, such as robots?What about the model’s identity?11RobotsDoes it matter who we are learning from?The ‘like-me’ hypothesis would make a clear prediction …Human > RobotThe aim of thispractical andResearch Reportis to test thisprediction12Experimental details• The data that will be used in your research reporthave already been collected.• The data will be made available to you via iLearn andat the end of this presentation, where you can copyor download the full dataset and then analyse it aspart of your research report.• For the remainder of the practical, you will:– have a go completing the task, so that it is familiar to you.– Find out more of the experimental and procedural details.13Basic experimental designHuman model vs Robot model14First, have a go at the task• Get a laptop and work in pairs or small groups.• One person physically practices, the others watch.• Place four fingers from your non-dominant hand(typically left), on the number keys 1-4.12 3 4 1 2 3415First, have a go at the task• One person physically practices. Using one finger perkey, hit the number keys in the following order, asfast and as accurately as possible (Repeat x 10).• Everyone else watch carefully.16Now, switch positions and test• Switch over: those who were watching have a go….• Repeat x 1017Now, try a difference sequence…• Now try a different, untrained sequence….• Repeat x 1018Repeat the process yourselves…• Now, write down your own sequences on a piece ofpaper or laptop and repeat the training procedure.• One person practices, the others watch.• Then one person who watched gets to testthemselves on trained and untrained sequences.19Task summary• Ok, so now you should have a good idea about thebasic task.• Physically practicing and watching others performsequences makes you faster and more accurate thanperforming untrained sequences.• But does learning differ if we manipulate the identityof the model and try to learn from a robot? Thecurrent practical addresses this question.20Independent VariableIndependent VariableType of training. Three levels:– Human, robot, no training.Repeated measures or within-participant design (i.e.,every participant is in every condition). For eachparticipant, a set of sequences are assigned to eachcondition. E.g.,Sequence 1 2 3 4 5 6 7 8 9 10 11 12Condition Human Robot No training21Dependent VariableDependent VariableSequence completion time (s). The time taken tocomplete 4 key-presses in a specified order.22ProcedureThe procedure involves two phases (training and test).Training Test– Sequences are observed beingperformed by two differentmodels– All sequences are performed– 4 human sequences– 4 robot sequences– 4 human sequences– 4 robot sequences– 4 untrained sequences23Data• Each participant completed a series of trials percondition in the test phase.• The mean execution time in these trials wascalculated to represent the average speed percondition. See the data below for the first 4participants (pID = participant ID).pID Human Robot Notraining1 1.93 1.26 2.472 1.25 1.68 1.033 1.02 1.67 24 0.67 0.89 1.6524Hypotheses• A hypothesis combines the IVs and the DVs into aprediction statement• A good hypothesis statement is an “if” “then”statement• The IV is the “if” and the DV is the “then” part• What are we manipulating (IV)?• What effect do we expect from that manipulation onthe DV?• Are we expecting a particular direction in thatrelationship?25Analyses• Your statistical training from Year 1 is sufficientto analyse these data.• Include:– Descriptive statistics (e.g., mean, SD per condition)– A plot of the main findings that illustrates the data– An inferential statistical test, which addresses your keyhypothesisHot tip: Keep it simple!!26Possible discussion topics• What were the effects of watching a human compared to a robot?• Was there an improved performance for the human compared to therobot condition? If so, can it be unequivocally attributed to theorisingfrom the “like-me” hypothesis?• How do the results potentially relate to the function of mirror neurons?• To what extent can we generalise from the findings with this task toother tasks? Can you think of tasks where a robot model might besuperior to a human model?• We used one type of robot, but how would the results relate to othertypes of robotic agent?• In the light of the data you have gathered, suggest ways in which otherpractical tasks might be organized for optimal learning.27The dataset• You can copy and paste the datadirectly into statistical analysissoftware, excel or whatever youwill find useful.• You can also download the datafrom iLearn.• You can download Stata usingthe following link:https://students.mq.edu.au/support/technology/software/statapID human robot no training1 1.93 1.26 2.472 1.25 1.68 1.033 1.02 1.67 24 0.67 0.89 1.655 1.23 1.63 1.976 1.04 2.51 2.367 0.87 1.92 0.688 0.99 1.85 1.419 1.09 1.45 1.1710 0.99 1.57 1.7911 1.42 1.35 2.212 1.05 1.63 2.3813 1.49 1.92 2.4114 1.88 1.09 1.9415 1.94 2.46 1.4816 0.93 0.97 1.7817 1.05 1.3 1.9718 0.71 1.28 1.2419 1.69 1.24 1.0120 0.37 1.68 1.521 1.3 1.6 1.7322 0.76 1.07 1.8523 0.67 1.41 1.2524 2.08 1.23 1.5325 1.21 1.49 1.4626 1.84 1.28 1.2527 1.03 1.46 2.0828 0.48 1.97 0.7929 0.42 2.3 2.1330 0.47 1.33 2.2331 0.95 0.62 1.0832 2.23 1.73 1.0733 1.63 1.84 1.2134 1.18 0.94 2.0235 1.05 1.72 1.7136 1.76 1.25 2.3437 2.85 0.8 1.3238 1.51 1.54 1.639 1.37 1.64 1.5640 1.16 1.41 1.6941 0.62 0.99 2.8442 1.41 1.54 1.8243 1.77 0.8 1.5244 1.19 1.07 1.7545 1.3 2.01 1.7746 1.25 1.09 2.1347 0.73 1.43 1.6348 1.47 2.37 1.1349 0.64 1.13 1.2350 1.06 2.05 2.9128Any questions?More information can be found on iLearn, including:• The slides from today’s practical.• A Research Report Assignment Sheet, whichsummarises the assignment and the experimentaldata.• Background reading articles.• An assessment rubric.