Mailbag #162
/
So, I'm considering doing some small, informal testing of different variations of the same trick for a few people, to find out which they prefer. But, like Natalie Imbruglia, I'm torn:
On the one hand, it'd be logistically easier to show them all the variations together, as a group.
On the other hand, to eliminate the possible primacy effect of people favouring whichever variation they see first, I should ideally do the test several times, with the variations in a different order each time.
Since you've done a bunch of focus group testing, I'd love to know whether you've generally found the primacy effect to be strong (in which case I should definitely do it multiple times), weak (in which case I could just do it once), or variable depending on the trick (e.g. if it contains a surprise)?—OM
In testing, you have to worry about both the primacy effect (which one they see first) and the recency effect (which one they see last). Both of these can affect the results and not addressing them would be, as Natalie Imbruglia said, a big mistake.
In our testing, we would let the variables we were testing determine if we could show them to the same person (or group) or if we needed different people for each variable.
So, for example, let's say the question was, "What do people like best? Card tricks, coin tricks, or mentalism?" These are different enough variables that we can show each person one of each, and just rotate the tricks with each person so no one trick is always first, middle, or last.
But if you're testing similar ideas/tricks, then you need to separate them completely. For example, "Is it more impossible if we frame this card trick as me reading their mind or them reading my mind?" You can't just do both tricks and then ask which they prefer. I mean, you can, but like Natalie Imbruglia said, you'd get the wrong impression.
That being said, we were trying to do testing on a broad scale with dozens or even hundreds of respondents. But I think a lot of these things you can test in very casual ways by just talking them out with a few friends and get decent enough information from that without even performing the trick. You can just talk through the variables—"Would you think it was more impossible if I picked a card and you guessed what it was? Or you freely named any card and it was the one card in my wallet?" This works as long as the differences are easily verbalized, not some subtle physical thing like a variation in a double turnover.
Testing out ideas is great, but the time and money it takes to do it “the right way” can be prohibitive. Talking through tricks with people is a worthwhile alternative to testing. And you don't have to concern yourself like we always tried to with things like primacy bias, recency bias, demand characteristics, fluency bias, contrast effect and other things that would make your head spin, like the titular character in the Neil Finn song, by way of Natalie Imbruglia, Pineapple Head.
I’m curious whether propless techniques such as hanging statements, making a statement and gauging reactions, and similar approaches work reliably in social environments. These techniques are often used in propless mentalism when the choices are narrowed down to two. —MG
In my experience, these techniques don't work great in social situations. At least not in ones where you know the other person.
Why? Because they're predicated on people acting in a generic way toward you. That's how people might act if they're audience members in a show or if they're strangers you've never spoken to. But that's not how people react when interacting with friends or even acquaintances.
For example, here's my conversation walking into a new cafe for the first time.
Me: Hello.
Counter-person: Hello. How are you today?
Me: Good. And you?
Counter-person: I'm good.
Me: I'll have a cold brew.
Counter-person: $5.19.
Me: Thanks.
Counter-person: Thank you.
It's a standard, generic interaction. And you could predict what each person will say next or how they'll react along the way.
But when I walked into this cafe a few moments ago, it went like this:
Me: Hello.
Sarah (barista): Oh god. What now.
Me: [hopping forward, step by step and making a farting sound each time my foot hits the ground] pfft, pfft, pfft.
Sarah: Actually, I was hoping you would come. I have a story for you. I had to call the cops on someone this morning.
If you tried to predict this interaction, you would be screwed.
Similarly, saying something like, "I want you to repeat your card over and over in your mind, like, 'Four of Spades. Four of Spades,'" and looking for the way they react can work when you're dealing with a stranger who reacts generically.
But friends and acquaintances might smile or be straight-faced or give a look that's contrary to what mentalism techniques would suggest, because they have a whole history and style of interaction with you that doesn't follow the rules of performer/audience.
That's why I prefer very concrete techniques that don't require me to pick up on a subtle look or reaction.
Your force bag conversation hearts trick is a great idea. I'll try it out with one change. Instead of them tossing it in your mouth, they will put the candy in their mouth, and you will determine the message by French kissing them. That could lead to a great Valentine's Day.—GT
Yes, this was the initial idea, but I don’t trust magicians to know when this is appropriate or not. Here’s the deal, if you’re already kissing this person outside of this trick, then feel free to perform it like this.