The other day I posted the reach of a BOOK BOMB, and then I got this unique, and interesting request from my favorite brain scientist (seriously, I’ve toured his lab. It would blow your mind)
Larry:
After seeing the success of the Book Bombs, I wonder if you would consider a “Survey Bomb” for me? We’ve had a really good response so far, but I’d like to give it just that added push to put it over the top.
Thanks for considering it!
Best,
-Speaker
Survey text below:
————————-
This is a survey for a science lab that is trying to categorize a set of images for some memory experiments. We want to be able to analyze our experimental data on the basis of the “category” of pictures used in the test, but before we can do that, we have to get a mostly objective categorization of the images we use, hence this survey: http://hampsonlab.org
The survey consists of a page of 25 pictures on a single page (we have 20 different, randomly presented pages, but most people complete just a single page), and there are checkboxes to choose up to 20 features or “categories” that you see in each picture. We know that there are pictures that are hard to categorize with the limited selection we provide – that’s *on purpose*, since the categories were chosen based on prior research. Just pick the *closest* category match. If you don’t like any of the choices, leave them blank.
This is not “dual use” data since we will not utilize the data for any purpose applicable to the people who take the survey. All responses are *anonymous* – we save only the counts for category responses per image – and that data is stored directly on the server. We’re trying to get a high-enough response that the results are statistically appropriate – to do so, we need well over 500 individuals to respond and complete one or more survey pages. So far, we have about 300-350 individual responses, completing 433 survey pages. I’ve asked Larry’s help because I know the Monster Hunter Nation is a powerful force for good! In this case, let’s make it a powerful force for Science!
Many thanks,
-Tedd Roberts, the Speaker to Lab Animals
So there you go. SCIENCE BOMB! 🙂
I took the survey. For some reason I felt like I was in Blade Runner
Yeah.
That spider with orange body and green legs who lived behind my window once and his huge egg… that picture was the worst one.
Done a few of them for Speaker! 🙂
Science will thank me someday.
My gratitude.
Speaker- You need, at minimum, another 200 plus… and you’ve decided to ask the MHN… oh, son, I hope you’re ready…
Done!
A lot of the pictures were missing the main color I saw. Took 2 pages of the survey, hope that helps.
Conned both my wife and grandson into doing a couple of surveys each. Gonna cost me doing the dishes and pizza.
Done!
Done!
Color palette is FAR too limited. Brown, grey, black, pink, etc. are needed. Make sure to let us know the final numbers 🙂
How do you know that’s not what they’re testing? They may be trying to figure out what basic colors people use to substitute for describing more subtle tints.
People who do polls in order to learn things (as opposed to people who do them to find support for a conclusion they’ve already chosen) can sometimes gain more insight from how people react to poorly worded questions than from answers to plain questions.
At my last job they did an annual employee engagement survey, and every year one of the questions was, “I have a best friend at work. 1) Strongly disagree 2) disagree 3) No opinion 4) Agree 5) Strongly agree.” My group (all design engineers) spent hours arguing over how poorly worded the question was, trying to guess what the pollsters really wanted to ask, and how the question should have been worded. Our manager finally agreed to ask the head of the polling committee about that question just to shut us up. After the survey was done he told us the answer: they had discovered that the productivity of any given department correlated directly with the response to that question as written. They didn’t give a damn about whether we had a friend at work or not, they just wanted to know if we thought we had one, according to however each of us defined what “friend at work” meant.
It sounds like we should assume that all surveys have purposes out of Kafka and refuse to answer them. And indeed, I would consider that approach if the survey came from my boss at work or the government. But if we’re willing to do the survey honestly, then that should include speaking up when the survey writer appears to have made a mistake or to be full of hooey.
At the company I work for, we have a system of testing to move from one job to another. many of the tests for better jobs have poorly worded or out of place questions just to make you think aren’t trained for them. the best example I could give was one for basic electronic troubleshooting that had a question about figuring coulombs in a circuit. I knew several people who saw that question on the first page and handed the test in without looking at the rest.
I can almost give you the names of the colors of the floss 😉
This is interesting. I am curious how the researcher, I believe Dr. Roberts, will account for issues that could come up from a large influx of surveys that may have a certain level of homogeneity in regards to the survey taker’s characteristics. If there is no cap on the number of participants, and if the previous book bomb is any indication, it is possible that the number of participants recruited here could easily become equal to or greater in number than what was already collected.
Just some thoughts. Keep in mind I am simply an amateur scientist.
I’m curious as to what you think the common characteristics of Larry’s readers are (other than they read things he’s written)
Its just a thought, but lets say hypothetically we are on average smarter than the general populace. Some research indicates that people who read for enjoyment are more intelligent than the average person. Intelligence affects multiple areas of cognitive and brain studies, whether it involves simply assisting with categorization for memory experiments or taking the experiment itself. Since Larry produces a good number of books and book bombs for other people, it seems fair to state that in general we, his fans, read more than the average person. If the correlation between intelligence and reading is accurate, a huge influx could potentially skew the part of the study we are assisting with, not necessarily to a huge extent, but still could.
Its also entirely possible that it may not skew anything important at all. Its something I started to think about when I do studies with people in regards to memory.
Many thanks to those who helped out with the survey, For those interested, we had 453 responses to the survey within 24 hours – we went from 433 to 886 total. The MHN is *AWESOME* and I thank you all for your support.
Regarding the most common comment – need for more color tags – the truth is, the *memory* doesn’t always work that way. In fact, what we have found is that only the most memorable *simple* characteristics are retained. Thus we used 6 main colors and “rainbow” and we used about a dozen high-level attributes for the content tags.
Again, this is awesome, and we are certainly now in a range of statistical validity – at least on par with any internet survey that does *not* seek demographics. And just as a reminder, we don’t keep *any* personal data on you, just counts of how may times the tags were chosen per images!
Did you run your “ad” over at the Bar, too? Or was that a lower response expectancy?
Anyway, I would love to see the statistical summary sometime. Particularly wondering how many (if any) saw the cactus blossom and checked “food” without a second thought.
I think the problem with your color tags answer is that most people categorize whites, grays, blacks and tans/browns as colors as well. The fact that something may not have any VIBRANT colors doesn’t mean (for most of us) that it is “colorless”, except perhaps in an purely relative sense. A picture of a cloudy blue sky is going to register as having TWO colors, blue and white (assuming fluffy white clouds and not angry gray ones). A picture of a cloudy gray sky over a extensive sand dunes will be gray and tan. And if there’s a single small bush in the corner of the pic, while green may BE there, it doesn’t qualify as “most noticeable.” Yes, I realize part of what you may be investigating is where the shift point between most noticeable and incidental is at.
Good luck with the research.