Investigating Inattentional Blindness

[EnglishRose98]

Thank you Science Buddies! The team you have is great!!

I have yet another question.... I'm just about ready to dive into the experimentation part of my project (I still need to accomplish some more research)
BUT, how does this look?

Question: “Are Females more perceptional and observational than males?”

Hypothesis: “If females are more perceptional and observational than males, there will be a higher percentage of females than males who notice any unusual incident(s) in the videos.”

Materials:
 two Video Recordings (they are similar, but do have their diferences)
 TV or computer
 At least 50 test subjects
 Parental Consent Forms
 Chart to record data

Brief Experiment and Research Overview:
This Behavioral & Social Science project is about inattenional blindness. It deals with whether people are able to focus on a particular object or action AND notice what is going around them. Though many of us don’t realize it, we all have the tendency, to some extent, to focus on only one thing at a time, and be completely oblivious of other things going on around us. This is how we were designed, and without this ability, our everyday encounters with life would often have a complete lack of order. "Only those items which I notice shape my mind - without selective interest, experience is utter chaos"; "The function of ignoring, of inattention, is as vital a factor in mental progress as the function of attention itself."(William James, 1890). In order to gather evidence that either supports or disproves my hypothesis, I will gather 50 subjects and play for each of them two short (approx. 4 minutes altogether) video recordings of two particular scenes. Afterwards, each subject will be asked a few simple questions related to what the subject has just watched.

Experiment Procedure:
1. Get permission from all parents.
2. Read to each subject the instructions of what to do for video A.
Instructions to read each subject:
Today I will be conducting a simple observational test. It is designed to highlight how males and females think. I will be showing you a video of six people passing a basketball. Three people are dressed in light clothes and they are passing a basketball between themselves. Your job in this test is to count the number of times they throw the basketball to each other. To make the task more challenging, there are three people dressed in dark clothes who are passing a basketball. Ignore these people and just concentrate on counting the number of passes that the people in light clothes catch.

3. Have all of the subjects watch video A and follow instructions.
4. Ask question #1a. How many times was the ball passed?
5. Read the instructions for video B:
Now, I will again be showing you a video of six people passing a basketball. The directions are the same. Just count how many times the people dressed in white shirts pass the ball.
6. Have all of the subjects watch video B.

7. Ask question # 1 b. How many times did the people in white shirts pass the ball to each other?
8. Ask the two questions below for both videos
Did you notice anything unusual?
What did you see? (To be asked if the subject answers “yes” to previous question)
9. Finally, compare results and draw a conclusion.


Note: Considering this is an experiment that involves human subjects, my actual first step will be receiving permission from each subject’s guardian or parent(s). I have created a form containing a brief experiment overview; I am going to have a parent of each of the subjects sign. I will also need each subject’s signature.

Once again, thank you!!!

[heatherL]

Hi EnglishRose,

You have put together a really great experimental design! I'm excited to see it coming together.

I think it looks really good, but I am wondering why you chose to use two different videos. How are the videos different from each other? How do you plan to analyze the data? For example, some people may perform better on the first video than the second, and vice versa. Since there may be differences within individuals between videos, how will you compare the different individuals to each other?

Heather

[EnglishRose98]

Hi,
I did consider a few of you questions also... I just felt it would be better to do two - wouldn't it be better to have more data and ways to test my hypothesis?

The first video will show three people in light clothes passing a basket ball, and three people dressed in darker clothing passing a basket ball as well.
In the middle of the video, a person dressed in a gorilla suit comes onto the scene, and then exits. The subjects supposedly will be so focused on counting how many times the people dressed in white pass the ball, that they won't notice the gorilla. After the video has been watched, I will only ask them how many times the ball in the specified group was passed.

Next,
I will show them another video. The instructions are the same, but the background setting is a little different. Six people are now on a stage; behind them is a red curtain, in the middle of the video, a "gorilla" comes out on stage and leaves, but also, a player dressed in dark clothes leaves the stage also AND the curtain changes color from red to gold.
After this final video has been played, I again ask them "How many times did the people dressed in white pass the ball?"

Then, for each video, I'll ask each subject "Did you notice anything unusual? If so, what did you see?"

I was also going to ask them at this stage if they had seen any of these videos before. What should I do if they were familiar with the videos and had watched similar ones? Surely, that would affect the results, wouldn't it?
How does that sound? The videos are online - I can post them. Should I just play the first video?

Thank you:).

[heatherL]

Hi EnglishRose,

Your videos both sound great! It is not always better to have more data if it will make things more confusing, but you can do the more complicated analysis if you're interested in it.

It comes down to what question you are trying to answer. If you are most interested in gender differences, I think using only one video will give you the clearest picture. If, however, you are also interested in whether people improve their ability to observe multiple things from the first to the second video, you can examine that as well.

I do think you will need to ask your subjects (after the experiment) whether they had seen the video previously. If the answer is yes, I recommend removing them from the data analysis.

When it comes time to analyze your data, you will have to take multiple approaches. You will need to see (a) whether people improved their observations between videos and (b) whether the responses varied by gender. To do this, the best statistical analysis is called a mixed linear model, which allows you to account for multiple independent variables (gender, video) and dependent variables (accuracy of count, observation of gorilla). You will need the help of a statistical program and/or a person with good statistical knowledge to analyze your data.

If you use only one video, you can do a simple t test (in Excel, for example) to compare males and females in their ability to notice the gorilla while trying to count the passes. You could also try to see whether counting accuracy affected their observation of the gorilla.

So it comes down to whether you want to learn some more advanced statistics while doing this project. I think it will be interesting enough to look at one video, and try to see the interaction of counting accuracy and peripheral observation, but it is really up to you! Let me know what you want to do, and I'll try my best to help you along the way.

Heather

[EnglishRose98]

Thank you so much; your answers are always so thorough and informative!
I think I'll just do one video.. Could you please advise me on which one??

Thank you in advance:).

[heatherL]

Hi EnglishRose,

I'm glad to help! I think the simplest way to go is to use only the first video. After the subjects watch the video, you will ask them how many times the people in light clothes passed the ball, and record their answers. Then you will ask them if they noticed anything unusual during the video, and ask them what it was if they answer yes. This will give you only two major dependent variables to follow: (1) count accuracy and (2) whether they noticed the gorilla.

The second video sounds really great, but it introduces multiple variables that could make it more difficult to analyze the data. In addition to count accuracy, you have three different unusual events that the subjects may or may not notice: (1) the "gorilla," (2) the person leaving, and (3) the curtain color. It would be interesting to see what people notice of those three events, but it might make the data analysis a little more complicated. That said, it would be very interesting to see whether there are gender differences in what people notice when they do (e.g., if males notice the gorilla and females notice the curtain color). So you could go for it if you are interested in the different ways that selective attention might manifest according to gender.

Sorry for the long answer, but in the end it is up to you. The first video is clearly the simplest route in terms of data analysis, and the second video will be more complicated but could answer additional questions. Ultimately, it depends on the questions that are most interesting to you. If your primary concern is gender differences in selective attention, go with the first video. If you also want to know what males and females notice, then use the second one.

Please keep me posted on your progress!

Heather

[EnglishRose98]

Hi again!

I have now come to the data analysis part.....
I don't quite know what to do because my data neither supported nor disproved my hypothesis.....
Here is the data I collected:

Of the 58 people I tested,
31 were girls
27 were boys

Saw gorilla: (Didn't see gorilla):
17 girls 14 girls
20 boys 7 boys



Each time, it seemed to depend on the individual being tested. With some people, I found myself very surprised they saw the gorilla, and with others, it was the opposite...
I didn't notice any real pattern of who did or didn't see the gorilla. As an example, there were two men that I tested who were 60 years old. One counted like 14 passes and missed the gorilla; the other counted seven (I'm a little suspicious!) and saw the gorilla. There were two girls that both counted 14 passes; they both were 13 years old, but one saw the gorilla and the other did not..
I'm also wondering if I spaced the ages too far out.... I had a range of 9 - 60!!!
Two people were 9
Two were 10
Two were 11
Seven were 12
Eleven were 13
Seven were 14
Four were 15
Four were16
One was 17
One was 18
Seventeen were above 20 [21-60]

I welcome all the input you have time to give!!
Unfortunately, I have to analyze ALL the data today!! I just don't think I can say that boys are more perceptional and observational than girls....Even though more boys saw the gorilla than boys, I would say that my data sample was too small to get to that conclusion....
Thank you in advance!

[EnglishRose98]

Oh, I also meant to ask about another site I thought Sceince BUddies reffered people to when they were trying to create graphs... I have searched the S.B. website, but cannot figure out where on earth I had seen it! Does such a thing exist on your website, or did I find that somewhere else??

Thanks!

[heatherL]

Hi EnglishRose,

I apologize for the delay in responding. It looks like you have very interesting results, and it's okay if your hypothesis was not supported.

One way to tell if the girls and boys were different is to do a t test (http://en.wikipedia.org/wiki/Student%27s_t-test). This is a simple statistical test that you can do in Microsoft Excel (with the Add-in called Analysis Toolpak). Organize your data by gender, with one column for girls and one column for boys. If someone saw the gorilla, he/she gets a "score" of 1. If someone did not see the gorilla, he/she gets a "score" of 0. Then you can compare the two groups to see whether they are different. I ran a quick test, and it does not look statistically different. However, there is another important factor: whether they counted right.

If someone counted the right number of passes AND saw the gorilla, that person gets a score of 2. If someone saw the gorilla but counted the wrong number of passes, you could give that person a score of 1. If someone counted the right number of passes but did not see the gorilla, that person would also score 1. If someone missed the gorilla and counted the wrong number of passes, that person gets a score of 0. You could try comparing people that way to see how gender and age affected their attention.

You may not have to use statistics (though it certainly impresses the judges at this level). You could visualize your data by making bar graphs. Again, I think it's important that you look at 4 groups: (1) people who saw everything, (2) people who missed the gorilla because they were busy counting, (3) people who saw the gorilla but messed up the counting, and (4) people who missed the gorilla and miscounted. You could compare these by gender and by age. The age is difficult to analyze, but doing a bar graph by age could be interesting.

Here is the Science Buddies information for data analysis and graphs: http://www.sciencebuddies.org/science-f ... ysis.shtml

Let me know if this makes sense, and definitely write again if you have more questions.

Heather