Implicit associations and brand preferences

The Implicit Association Test (IAT) is an intriguing thing. You might have heard of it, since Harvard has been running a pretty high-profile project using variations of the test for a while now, and the assortment of results certainly gives some food for thought from both psychological and sociological perspectives. The IAT is designed to test the associations that you make between things. These associations are implicit because they can even run contrary to your explicit opinions – they’re associations that your brain makes that you might not even be consciously aware of.

The overall idea is that when you have to process information about things that you implicitly don’t associate with each other, you’ll be slower at it than when you process information about things you do implicitly associate with each other. For example, you might be slower to respond to the word combination of “science/female” than to “science/male” because you might have the implicit association that science is more of a male thing. This doesn’t mean you’re a sexist and hold deeply ingrained discriminatory beliefs – it could easily be because of cultural and social influence that you’ve come to associate “science” with “male” more so than with “female”. You can be the world’s most super-informed, acutely passionate third-wave feminist and still have that implicit bias against “science/female”.

Anyway, the point is that using the IAT is a good way to understand what might be going on in people’s brains in a way that you couldn’t achieve just by asking “So how do you feel about this?”. (Disclaimer: there are plenty of caveats that go with using the IAT, which are more or less summed up here.)

As you might imagine, the test is pretty interesting to use in the consumer setting and it can present a novel view of how products or marketing or whatever might affect your implicit attitudes towards brands, behaviours, etc. That is something that Ratliff et al. investigated in their paper, ‘Does one bad apple(juice) spoil the bunch? Implicit attitudes toward one product transfer to other products by the same brand’. Ratliff and colleagues wanted to examine people’s attitudes towards products from the same brand, and how your attitude towards one product from a particular brand spills over to affect your attitude towards other products from that brand. In the first part of the study, the researchers wanted to check that you get this kind of spill-over effect, where having a positive attitude about one of the brand’s items positively affects your appraisal of other items from the same brand.

This was indeed the case – in the study, if people had a positive attitude towards a moisturising lotion of Brand A and a negative attitude towards a moisturing lotion of Brand B, and were then presented with totally neutral and equivalent descriptions of a deodorant from Brand A and a deodorant from Brand B, they of course preferred the Brand A deodorant, even though the information they had about the two deodorants was ostensibly identical. In this case there was implicit and explicit attitude transfer, with participants preferring the deodorant of the positively perceived brand both when tested with the IAT and when filling out a scale to rate how they felt about the deodorants.

But of course, we wouldn’t be evaluating that deodorant in a vacuum, so to speak – we usually have some degree of positive or negative information about a product already, like whether it contains ingredients we’d prefer to avoid or whether its quality doesn’t seem all that great or whether it ticks all the boxes we want it to. So what happens when you like one product from a brand, and then have to evaluate another product, which either seems good or not so good? Does that pre-existing positive attitude about a brand’s product influence your evaluation of that other product? Does liking that one other item from the brand mean that you perhaps evaluate a bad product more positively than you otherwise would have?

To start with, participants read information about taste tests and were thus induced to have a positive attitude towards Brand A of apple juice and a negative attitude towards Brand B of apple juice. Participants then evaluated orange juices by these brands based on written descriptions of taste tests, the results of those taste tests being either positive or negative.

The results? When attitude to Brand A apple juice was positive, and then positive information about Brand A orange juice was presented (and negative information about Brand B orange juice was presented), participants explicitly preferred Brand A orange juice over Brand B. When attitude to Brand A apple juice was positive, but negative information about Brand A orange juice was presented (and positive information about Brand B orange juice was presented), participants explicitly preferred Brand B orange juice over Brand A. So even though Brand A’s apple juice was great, participants didn’t let it colour their appraisal of how Brand A’s orange juice was not good.

The interesting thing here is that these explicit preferences do not match up with participants’ implicit preferences. When attitude to Brand A apple juice was positive, and Brand A orange juice was presented as kind of bad (and Brand B orange juice as kind of good), people still implicitly preferred Brand A orange juice. Their attitude towards the apple juice was colouring their appraisal of the orange juice by the same brand, even though they seemed to know at a more explicit level that they shouldn’t let their judgement be swayed like that.

That’s the power of liking a brand’s product – to some extent, at some level below conscious awareness, your brain is still saying that the brand’s other products must be good, even in the face of negative information about those other products.

We can’t know from this study whether these implicit biases have particular consequences or behavioural outcomes (e.g. does having this implicit bias mean that people are more likely to actually buy the poorer product, or would the explicit attitudes prevail when it comes to making that purchase decision?) and there are plenty of other factors to consider when interpreting the results (e.g. the results might very likely be different if the participants formed their own attitudes rather than getting them by reading about other people’s opinions from taste tests). But it’s incredibly interesting to consider the extent to which we could be unconsciously influenced by brand preferences, regardless of other information that might provide a better basis for our judgement.

8 Thoughts on “Implicit associations and brand preferences

  1. I love this. My current research is also rooted in the IAT principle (related to cognitive capacity, peripheral processing, etc.) but I have never thought about how much it also applies to my shopping behaviour. I do think that most of the time implicit associations improve our decisions. It is nearly impossible to objectively consider every little piece of information there is about two choices, so transferring what we know about one product to another from the same brand, even if that happens implicitly, is our best bet.

    • Jess on April 2, 2013 at 6:47 pm said:

      I think it’s really interesting to consider how those sorts of implicit associations affect our decisions, because it’s kind of like adding a cognitive layer to the physiological correlates of intuition – like the whole idea that decisions based on “gut feelings” are often better than decisions based on too much thinking and analysis. So the autonomic nervous system plays a part in generating that gut feeling through some sort of low-level appraisal of the information, but I wonder to what extent implicit associations also contribution to that sort of intuition. It would be interesting to try and see if those effects are separable!

  2. A cool way to show the powers and the limitations of branding. I think the idea of threshold in your own personal shirt example is how there is a certain threshold perhaps that a product must cross in order to win brand loyalty. It might not be the best in absolute terms, but since the shirts from Claudie Pierlot are good enough for you, there is little incentive for you to seek something totally new (and which can turn out either very good or very bad).

    I also wonder whether this can be tied in with more economic concepts of risk adverse and risk loving individuals?

    • Jess on April 2, 2013 at 6:59 pm said:

      I will definitely look up some papers about risk aversion and risk taking as factors in consumer behaviour – I’ve seen the topic come up quite a lot when I’ve been browsing databases and journals, but I can’t recall specific studies’ results off the top of my head. But yeah, it’s definitely another factor to consider, so the apple/orange juice study results could potentially be regressed to participants’ scores on risk aversion and that might break the results down into even finer detail.

  3. I wonder if this has anything to do with familiarity: it feels safer and more comforting to gravitate around familiar brands than take the “risk” to test a new one, even if objectively, there might be better stuff out there.

    And then, like Joy says, how far does a disappointment in a particular brand need to go to cut that implicit association? You mention that out of 2 Claudie Pierlot shirts, one of them is satisfying, but not the other one, yet you’d still buy a new shirt. I used to be a customer of Comptoir des Cotonniers for years, and probably had this implicit associations with new clothes from them based on the good experience I had on their first products. But over the years, a limit has been reached and when holes started to from in 60€ TShirts after less than a year, I suddenly stopped buying ther altogether. But it tool at least 2 to 3 seasons to sink in…

    • Jess on April 2, 2013 at 6:53 pm said:

      Ah, true, it would be very interesting to see how long that positive attitude can persist in the face of negative information. I would guess not long, since the actual quantitative results of the study showed that people only just preferred Brand A’s orange juice to Brand B’s, when Brand A’s was presented with negative information following a positive attitude towards Brand A’s apple juice. But yeah, you could just extend the study and after participants have evaluated the orange juice, they could then evaluate another juice with a negative description, and another juice with a negative description, and you could graph how perhaps the preference decreases and/or disappears over time.

  4. I think of going back to brands I had a good experience with as a shopping shortcut; it saves me so much time and effort if I know I can go back to the same place for quality items. But I’ve become more wary of that after experiences similar to yours with your shirts. I would think with fashion people remain “blind” to this longer than they would for say food, since people flock around a brand for reasons other than quality.

    • Jess on April 3, 2013 at 6:33 pm said:

      Oh yeah, it totally makes sense to seek out a brand you’ve had a good experience with in the past – it would be crazy not to. It’s just interesting that, when people have absolutely nothing to go on when judging an item except negative descriptions of it, they still prefer it because of one single previous preference for something else by the same brand. However, that’s a pretty unrealistic situation, evaluating an item on such limited information based on complete strangers’ opinions.

      And yeah, I agree that people would stay loyal to fashion brands for longer than food brands, given how much fashion brands seem to speak to people and their personalities and identities. I think people could have a bad initial experience with a fashion brand and still prefer it over other brands in the future, just because the brand means something very specific to them.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Post Navigation