Archive for the Uncategorized Category

The use of pictures in qualitative research

Posted in Uncategorized with tags , , , on September 2, 2009 by Cristi Popa

collage In qualitative research we tend to rely to some extent on images – may them be pictures or movies. Why? In no specific order:

• To impress clients. In this strategy, clients are seen as less than knowledgeable in research, they have their own “superficial” agenda (increasing sales, confirming their own hunches and thus get a raise etc.) and the attention span of a 10 year old (they are too tired/ bored/ lazy to follow an interview/ IDI so you need to “spice” it up a little. So you throw in a couple of collages/ picture sorting exercises.

• To have a pretext for starting/ animating a conversation. Pictures can be used either at the beginning of a group discussion (let’s say “choose from this stack of photos one that represents you and explain to the group the reasons for which you choose that picture”) or towards its end as an “energizing” exercise if people start to get bored/ tired.

• To elicit some data that otherwise (through discussion) would be overseen. In this strategy the assumption is that there are some ideas, feelings, emotions that are too subtle, too sensitive or evasive to be captured in words. Pictures come in as ways to get more information from respondents. So one can ask respondents to take some photos of let’s say their workplace if that’s our focus, or the places where they hang out etc. Or they can be asked to make/ cut out pictures that represent their position on different matters of interest: pictures that reflect their feelings when waiting to be checked in on airports, or which express how they feel when they are discriminated etc. Pictures thus selected serve as a baseline for the further discussion in which respondents account for the meanings they embed in those pictures. So, for example let’s say we are exploring the relationship between consumers and let’s say a refreshing drink brand. We select 10 respondents and ask each of the to take/ print from the web/ cut off from magazines 3 pictures that best describe their feelings towards that brand. Then in a one-on-one interview we take the photos the respondent brings and we explore the meanings invested in them. Let’s say that one picture shows an arm with a syringe in it. This might go like this: “Why this picture? Well it depicts what I’m feeling when I finish a bottle of X. I feel guilty because I realize that I’m poisoning myself. The sugar, all the additives are bad for me. But, still it’s like I’m hooked on it. I cannot go to sleep until I’ve finished the bottle….”. And then the moderator can go: “Why do you think it is bad for you?/ What do you mean by hooked” etc. and peel off the meanings of that photo which actually is an account of respondent’s relationship with the product/ brand/ etc. The same can be the routine for collages made in the middle of the interview or other photo exercises.

• To provide the researcher and finally the client with some peaks into respondents’ lives (bits of reality to either illustrate or to support the research findings). In here we already assume we have the meaning independent of the pictures (from the use of other methods) but we use pictures to support our findings, to give the report a more “life like” effect. So if we are talking about teenagers’ habits in engaging technology, after doing some interviews, after spending some time with them in their homes to see how they go about the subject, we also take some pictures of them listening to music, using a computer etc, to insert in the report. Nevertheless, at the data level they are worthless as they add no new knowledge to our account of the subject.

• To “test” them. This is the case for evaluating commercials, posters etc. In this case what we actually do is to check whether the respondents decode the stimuli (picture/ film/ text) as intended by the producers of those stimuli. In other words, we check if they adhere to the preferred reading of those pictures as it was formulated by the producers. This goes in the lines of: “What did you understand from this clip? Well it is about two guys competing for the attention of an attractive young girl by offering her beverage X. It tells us that if you want to get girls you need to invest in this drink…”. And let’s presume that this is not the intended manner to read this message. Then the researcher will state that this is not working and that the commercial is not in line with the associations consumers generally make with a beverage and specifically with the X brand.

So one can conclude that we do consider that a picture is worth a thousand words. Still, those words belong to those that invest meaning in them – respondents. To us, those pictures are mute until the respondents speak for them.
Moreover, one also could think that we can actually do without pictures for the first four bullets. We can impress clients by using other means than pictures (maybe some cool fancy software or some good wine in the viewing room?). Also, we can find other pretexts to spark up a conversation. And we can leave aside the illustrative role pictures might have in relation to our findings. Also, the disclosure pictures might facilitate can be imagined to be obtainable through the use of other objects: written accounts of respondents, drawings, music etc. I for one can imagine asking a respondent to bring some music that best describes his feelings towards some minority: “So why did you choose this song? Well it starts very powerful and brutal and reminds me of how we talk about a member of that minority when we first meet him/ her. But then the music goes softer, becomes more gentle and harmonious and that reminds me of all those people whom I got to know a little better and with whom I coexist in harmony although they are different…”.

So are pictures out of the picture? Not by far. And Gillian Rose in “Visual Methodologies – An Introduction to the Interpretation of Visual Materials” gives some fascinating accounts of what we can do with pictures starting from the evidence that our world becomes more and more visual.

For one, we can actually look at them. We are in the habit of analyzing just one site of meaning making: the audience, the meanings respondents derive from pictures. But as Rose points out, there are two other important sites of meaning: that of the production of that image and that of the image itself. And the methodologies implied by these other sites range from content analysis to semiotics, psychoanalytic approaches or discourse analysis.

Let’s return to the beverage X. Remember that we’ve asked people to bring us images of their relationship with that brand. Maybe that if we take a good look at those pictures we could see that all the images contained a reference to the body (lips, arms, stomachs etc.). Or that in most images actually showing the action of drinking there was always only one person. Maybe if we go a step forward we’ll realize that this drink is associated with guilt and shame and thus is not seen as suitable for drinking with friends or family. Or maybe we’ll see that it is more associated with different moments of the day. Or we could take a number of influential magazines and look at how they are portraying beverages in general. Do they emphasize the thirst element? Or the health? Are they showing mostly the product or its supposed beneficiary? Are they colored or black and white? Or we can make a little trip in history and see what refreshment meant 20, 30, 50 years ago. How were these products depicted or advertised in those days? Has anything changed? What? Why? Or when we analyze a commercial maybe we can go beyond checking the boxes provided by the agency and actually understand the embedded meanings from that ad, confront them against the general symbols used in this kind of ads or generally in relation to our product/ brand/ habit.

These are just some of the questions Gillian Rose’s book can shed some light on.


How Customers Think. Essential Insights into the Mind of the Market – Gerald Zaltman

Posted in Uncategorized with tags , , , , on February 10, 2009 by Cristi Popa

Some time back, Brian Baumal draw my attention towards this book by Gerald Zaltman. I never got to thank him so here it is: thank you Brian.

I remember that when I first read the book, it really got my hopes high. I was really enthusiastic for more than 200 pages. Then it just let me down.

A couple of weeks back I had the surprise of finding the same book, translated into Romanian (Cum gandesc consumatorii – Aspecte esentiale pentru studiile de piata, Polirom, 2007). I didn’t quite remembered why it let me down the first time I read it. So, I bought it, hoping that now, years later, after gathering more experience in research, I could properly understand it. So I read it again and of course remembered what disappointed me about this book.

Don’t get me wrong. The book makes for a good reading and it does get you thinking. Only for this reason and you should try it.

What bothers me most is that it just doesn’t keep its promise. The book starts taking a look at some of the newest (the book is published in 2003) advancements in neurology, psychology etc. and comes to a theory of what mind is, how thoughts are created, the role of emotion in this process etc. (nothing revolutionary here. More like a collection of already accepted facts about mind). Then it goes on saying that traditional qualitative research (mainly focus groups) does not stand a chance in uncovering the real motivations of the customers. So far, so good. But the whole time you’re reading you get this feeling that the author has a solution for all this. So you wait and wait and get your hopes high. You swallow the lack of structure, the vagueness of his examples and wait for his solution. And then the solution comes: metaphor elicitation – in other words an in depth interview in which the respondent is asked to bring with him a picture that best illustrates his feeling toward a brand, social problem etc.

In this point I thought: “but we already do this”. We also make collages, mind maps, diaries, ladders etc. And they’ve been around for years. I’m not saying that this is a bad method. But is this the promised salvation? The only way through which you will uncover the truth, the alpha and omega of qualitative research? Not really. It’s like in this episode of South Park when God comes to earth in a big, blinding light, with angel’s music on the background but he takes the shape of a drooling hypo-cat thing – the picture at the beginning ( min. 19:00).
south park

If you had a chance would you have tested the Obama idea in a focus group?

Posted in Uncategorized with tags , , , , , on January 27, 2009 by Cristi Popa

While we all hold our breath until someone will tell us why on earth qualitative research still exists, let’s take a look at some really funny clips about qualitative research (focus groups actually) and idea testing (may it be communication or product). The first is a parody of a prehistorically focus group on a “new idea”: fire. You must know it.

The second one is something I accidently came across here. The context is this: in 1984 Apple releases this spot:

It becomes iconic shortly after. 23 years later some clever advertiser does this clip in order to promote some advertising industry event. He makes storyboards with VO after the Apple ad and tests it in a focus group. The results are disastrous as you can see bellow.

Let’s set aside the fact that he was testing the ad with consumers from another generation so there was no chance to decode the ad as someone 25 years ago would do or for that matter let’s also ignore the fact that focus groups were not around 10 thousand years ago. Basically, both of them are saying the same thing: focus groups are not the indicated tools for testing innovative ideas or products. This idea is widely spread especially in advertising agencies. And I totally agree with it. You cannot test in a focus group too innovative stuff. It is only human for respondents to reject novelty because they get scared. Remember Galileo Galilei and the way that focus group ended. Still, how many of the commercials, products and ideas out there are that innovative? Not too many. And here’s a thought: was the Obama idea tested in groups? If so, with what results?

Quit bitching and moaning over qualitative research

Posted in Uncategorized with tags , , on January 26, 2009 by Cristi Popa

It became almost fashionable to criticize qualitative market research. The charges are various and range from the old count of lacking mathematical proof to the newer, hip ones – not thinking out of the box, not keeping up with the real world’s developments, generally not being cool and interesting enough. All these and much more are real to some extent. Still, it is a multibillion business each year, so one would think that there should be something good about qualitative research. But what are those things? So, the question is: why do you still buy it, read it, use it?

Why shouldn’t little boy-moderators play with dolls and little girl moderators with toy cars?

Posted in Uncategorized with tags , , , , on January 19, 2009 by Cristi Popa

13Bias is the Boogie man of qualitative research. It lurks everywhere around us: recruitment, stimuli, but, more importantly, moderation. Moderation can bias the results of research big time. Questions like “You like this, don’t you?”, remarks like “of course by this you mean …” are textbook examples of don’ts. But moderation bias is more insidious than that. It can originate in the tone of voice, dressing style, age, gender, body language etc. So, the little boys and girls moderators are careful to wear adequate clothes (resembling in style and price with those of the respondents), to have a similar phone, similar tastes (drink what they drink, smoke if they smoke, eat if they do), similar type of humour etc. before starting to play. Let’s say that the research agency has an IDI with a 35 y.o. housewife, mother of two, low to medium income, user of medium tier shampoos. Will it send the 25 y.o. little boy, a big fan of punk music and user only of premium shampoos which he “barrows” from his girlfriend? Or the 32 y.o. female that has a child and does her own shampoo shopping?

For the sake of the argument, let’s say that the boy is the only one available. So, the agency sends him. In this case, he will say to the respondent that he’s 28 y.o. so to narrow the gap, wind down his hair style, wear a shirt, borrow a low end Nokia from a colleague, say that he has a little nephew and generally lying his pants off to get the respondent think he’s her best friend. And, in the end, the moderator has to be a respondent’s best friend, so he has to be as similar as possible. But is that enough? Not really, because when getting to the part of usage habits, the respondent gets embarrassed to talk about how she’s getting naked in the tub. And so, there is a last thing the male moderator has to change. Fortunately for all you male moderators out there, it’s no need to. And here is why:

1.     First of all, respondents are usually not retarded. They will find it very odd that you “by accident” are so similar to them. And pretty soon they will start to challenge you.

2.     Moderating is hard enough without having to watch everything you do and say in order not to blow your cover. Giving attention not to reveal your “secret identity” will take a lot of your energy from the actual moderating process.

3.     And, most importantly, if the ideal would be to have a perfect identical moderator for each respondent, then why do we even bother to interview her. The moderator can meditate a little about the discussion topic and directly write the report.

In order to find another trap of this line of thought, let’s go back to the research agency and send the female moderator to do the interview. She is pretty similar to the respondent, as you might remember. So she has no reason to lose any energy by lying. She is practically her best friend. But, wait a minute, she still doesn’t find out what happens in that tub. What is happening? Well, the respondent is acting as a friend. And between such good friends that have so much in common there is no need for too many words. Certain complicity is born. You can spot it in phrases like “you know…when you enter the tub…” and then a smile. And the female moderator would say “well people are different so please tell me from your point of view what happens”. And the respondent replies “well it’s obvious. Aren’t you a woman too?” and so forth. The harder the moderator presses in this point, the more likely it is for the respondent to get annoyed and say anything just to get off the hook. The friendship is gone. And so is the interview.

So, what if we would drop all this similarity ideal and go for alterity? What if we would say to the respondents:  “I am different from you. I am a young man trying to understand how you wash your hair. I have no clue of how you do it, so please explain it to me in great detail. Being such a stranger to the subject will force me to ask you a tone of what may seem stupid questions, but please bear with me and make me understand”. I have tried it and it works. You wouldn’t believe the openness you are treated with after they understand your position. And that is why I know what happens in that tub.

In the end, here is a thought for all the girls and boys moderators: let’s just admit that it is impossible not to bias the data. Being alive means that we constantly interact with everything around us and by doing so we alter those things. There is no way to control all the variables coming into play in any interaction, may it be qualitative research or going to a movie. In fact, all we can do is to be aware of this and try as much as possible to take it into consideration in our analysis.