• The Listener
  • North & South
  • Noted
  • RNZ

Observational learning and the Bobo doll experiment

Observational learning changes our behaviour – and not always in a good way.

The Bobo doll experiment

Despite my best intentions to have my column ready well ahead this week, it hasn’t happened, and I’m procrastinating. Again. So, when eight-year-old Isaac says, “Can I look up Gangnam Style on YouTube? It’s gone viral you know”, it sounds like the perfect way to waste a few minutes. And, boy, do I regret it now. For those who don’t know, Gangnam Style is a “song” (yes, I mean the scare marks); it’s this summer’s catchy Macarena-pretender by Korean rapper Psy, and it comes with a particularly silly dance routine – that Wikipedia tells me will burn up to 400 calories an hour. It’s not long before Isaac is chanting “Oppan Gangnam Style” and mimicking riding a “sexy horsey”. And much as I cringe at the knowledge that this will define the next few weeks, he’s doing exactly what he and the rest of us are good at – observational learning, an idea most associated with the name of Albert Bandura.

Now, the idea that people (and particularly children) learn through observation is a bit of a duh moment, but in 1961, when Bandura did this work, the dominant idea about how we learn was based on behaviourism. Among the behaviourist tenets was the fundamental idea that our environment controls our behaviour, and learning occurs through reinforcement – put simply, being rewarded for something makes us more likely to do it again. Bandura thought this oversimplified matters, because it wasn’t always apparent what the direct reinforcement might be in all situations. Ultimately, Bandura developed his own theory in which he claimed that for learning to occur, people need to attend to what’s going on around them, that reinforcement isn’t just external but can be internal (feeling pride, shame, etc) and therefore psychological states are important, and finally that learning need not always result in actual change to behaviour.

But it’s not so much the idea itself as the manner that Bandura demonstrated it that’s particularly cool. This is the famous Bobo doll experiment. In this simple experiment, children (boys and girls, but always on their own) were taken into a playroom by the experimenter, and another adult introduced. The adult confederate proceeded to either play non-aggressively with the toys or metaphorically beat the stuffing out of a “Bobo” doll (an inflatable doll that bounces back upright after being knocked down). With a hammer. Unsurprisingly, when the children were finally placed in a room with a bunch of toys they were allowed to play with, they were more likely to play non-aggressively if they’d observed a non-aggressive role model, or beat the stuffing out of Bobo if they’d seen that done.

It’s also worth noting that not only were boys more aggressive than girls, but children were more likely to behave aggressively if their role model was male (even if the male role model didn’t behave aggressively himself). Bandura and colleagues argued this reflects that aggression is a more typically male characteristic. One early critique was that we can’t conclude viewing violent behaviour towards a doll will translate into violent behaviour towards people. I feel for the researcher in the clown suit in a subsequent replication – sadly, watching a role model whack Bobo did translate into clown-directed violence. As people who’ve tried out a career as a party clown will no doubt attest.

This is a big deal, and it kicked off a bunch of studies investigating the effects of things like TV violence on behaviour. To boil down thousands of studies, the overall conclusion is that exposure to media-portrayed violence causes aggression in some (but not all) people. This is true, even after we take into account the obvious fact that people who tend towards aggression also tend to watch more violent TV. No Bobo dolls were harmed in the preparation of this column.