(shameless self promotion of my research methods book)
I want you to experience a real psychological experiment, that also has some really interesting findings.
You can login as guest and it will take you to several options to choose from, or you can register for some of the other test options that require you to provide your email. This is a legit scientific site, so your email won't start getting spammed or anything like that.
I don't want to tell you too much to begin with...but basically these tests evaluate stereotypes that you may not realize you have. When you are done, there are informational pages you can click to to learn more.
Please take at least one of the tests. Describe your experience (what did it feel like taking the test). Were you surprised by the results (you do not need to discuss the topic of your test if you do not want to).
What aspects of research methods do you now notice from having completed the test? (like types of variables, control issues, reliability, validity, etc)
List your terms at the bottom.
The first test that I took was the gender-career IAT. In this test, I had to sort male names from female names and career words from family words. As time went on, they switched the sides that the words should be sorted into and also combined categories. It became harder because it switched from sorting female names and family on one side and male names and careers on the other to female names and careers on one side to male names and family words on the other. It was very difficult for me to adjust to this change. After the test, I had to answer a series of questions about who my primary caregiver was growing up and whether my parents worked while I was being raised. My results ended up saying that I had a “moderate automatic association of male with career and female with family.” My results were consistent with 32% of people in the study, the largest section of the group.
I agree with the results I got from this study for the most part. I would say that because of the way I was raised and because of the society in which I was raised, I do have an automatic association with females and family and consequently between males and career. The one reason that I am skeptical about the results of the study has to do with the fact that it relies heavily on the flexibility of the brain. Because the independent variable changed was the side that the categories were placed on as well as which other category they were grouped with, it made it difficult for my brain to keep track of which side I was supposed to sort words to. This made me wonder if this study really had proper causation for the claims it was making. I think that some of the reason that I often mistakenly placed words into incorrect categories was not because of the associations with the words but because I could not remember which side the category was located. For the most part, however, I thought that the study did a great job, especially in its follow up questions, in collecting data that might suggest why participants made certain connections between words.
Another interesting thing that I found was that the tests seemed to come up with stronger results when I had preconceived, conscious ideas about the subjects. For instance, when I took the Presidential Popularity IAT, I came out with only a slight automatic preference for Barack Obama. To me, this seemed to be because I had no particular knowledge about the difference in Thomas Jefferson’s policies and Barack Obama’s policies that would make me strongly favor one over the other. This made me believe more firmly that this test is designed to bring out preconceived ideas that you have about the subjects and not to necessarily discover something entirely new about yourself. This could affect the validity of the study because the researchers need to look more closely at the people in their sample to see what their connection with the subjects are before the test. If the participants answer honestly on the questionnaire after the test, the researchers will hopefully be able to eliminate this area of inaccuracy.
Vocabulary: Independent variable, causation, sample
It’s rather hard to describe how I felt while taking this test/ study, but I’ll try. I think at first I expected the stereotypes to be much less basic, and I expected them to be stereotypes that you would more often encounter in the real world, such as racism, so I may have gone into this with an odd mindset. At first the study was really odd because I was trying to determine what I was supposed to be doing and I was really disoriented. When I finally understood what was happening, with the association and the like, I took it to be a little less serious because the test wasn’t going to tell me how I stereotype people in the real world, but it instead had a smaller scale. After I realized this, I tried to just take the study at face value, and be unbiased. I wanted to beat the stereotypes, I think. After getting to the word association part of the study, I got a little less engaged, because it kind of told you when you were wrong, and I kind of wonder if that was their intention in creating this study, or if it was just a side effect that the creators of the study didn’t expect. I wasn’t really surprised by the test results, because once I knew that the test was about stereotypes, I kind of knew what the results would be based on the answers I gave. I was told I was only slightly biased when it came to stereotypes between the two people I was evaluating for character. I had kind of let the survey lead me into assuming things about the individuals based on their pictures and the preferences that were described in regards to each person, and I admit, I kind of fell into the trap of stereotyping the people based on the fact that one didn’t like puppies and kittens, while the other didn’t like bugs. One seemed more rational than the other, so I automatically liked him better.
The study did keep the test variables very stable, as it kept the same two people as the focus of the study, and only those two people. It also kept consistency with the likes and dislikes of each of those people, which made it easier to form a cohesive opinion. I’m not sure if the study was completely reliable, because I feel that if you tell someone that a test is going to determine how they stereotype others, then it will cause the person to attempt to prove that they are not stereotyping anyone, because stereotypes are associated with being a bad person. If the person finds out ahead of time what the study is on (much like I did) they will feel compelled to try to “beat” the study, which could lead to inconclusive results. Overall, though, I feel this is a minor concern, and the overall experiment has much reliability to it, and its results probably say something about all people when taken as a whole.
test variables
consistency
reliable
results
The first test I took was the gender-career IAT. In this test, I had to sort male and female names from career words and family. They would put male and career on one side of the board and female and family on the other and after a while they would switch their independent variables: the genders. As time went on it became harder because they had switched the groups and it was difficult for me to wrap my brain around it for a bit. After the test, I had to answer a series of questions about my caregiver and if they worked and what sort of work did they do while I was being raised. My results ended up saying I had a “slight association of Male with Career and Female with Family compared to Female with Career and Male with Family.” My results were consistent with 20% of people in the study.
I would have to say I disagree with the results I got from the study. I would say that I have an equal association with females and males with family and career. Maybe even more of a female and career association. My family life was different growing up then they seem to be supposing. But both my parents are workaholics, something I have long grown used to, but my parents make time for family. I think because the results of the study rely so heavily on the flexibility of the brain that it is not as accurate as it could be. This makes me wonder if they have proper causation for the claims they are making. Overall I think this experiment has less reliability to it, and they may take something else into account with this test.
Another test I took was the Religion IAT. In this test, I had to sort a specific religion and good words to the right while shifting the other religions and bad words to the left. This got to be confusing at times because sometimes they would repeat the same religion and sometimes they would switch it with another religion. Before the test it asked me to put down what religions I had already switched in my life and how important they were to me. After the test, I received my implicit preference scale. It seems after taking this Brief IAT I don’t really have a preference because every religion ranked in the middle of my preference scale.
As it is I would have to say I agree with the results I got from this study. I do not really have a preference for religion. I do not really go to church and have not read the bible and my family is not religious for the most part. I have been to many different ceremonies for different religions for my mom’s friends that I feel like I can say that I have a greater appreciation for those religion. All this study did was prove to myself what I thought I already knew which is great to know that I was right in this case.
Terms: results, causation, consistency, independent variable
Of the three tests that I took, the one I found most fascinating was the one pertaining to gender roles. This test included matching names with the corresponding titles such as Female, Male, Liberal Arts, or Sciences in a timely manner. While taking this test I felt like the study was checking my reflexes on how fast I answered a question and whether I got the answer correct. I later found out that I was correct, in part; this was exactly what they were doing in order to conclude a possible stereotype that I could have. In this way, I believe that this study did an excellent job. In one sense, however, there were sometimes that I messed up just because I had gotten into a rhythm not as much that I really believed that the particular topic went into that category. Overall, I do believe this study did a very good job of being unbiased or trying to sway the participants. This was a major topic in Chapter 2 that was a challenge for scientists to do and in this case I believe these scientists did a good job. Not only did they not include bias they incorporated the participant bias into their study.
This test was fascinating because it showed me more about participation bias and how crucial that is to eliminate as much of as possible or incorporate it in some way. For example, the result of the gender roles test for me was that I may have a slight association of Female with Science and Male with Liberal Arts compared to Male with Science and Female with Liberal Arts. I was part of 6% that have been given this conclusion. This could have been attributed to participant bias, considering, I am a biology major. This study really opened my eyes as to how big a part biases can be whether unintentionally set by the scientist or the participant.
On another note, it was interesting to me that these studies choose an inductive approach to gathering data. They gathered observations from specific adjectives from a broader topic and took that data to apply it to an even broader and more generalized conclusion. This was something that was brought up in chapter 2 and is seen in action here.
Since all of these tests were in survey form this has many pros and cons. For example, there was not as much researcher bias because I was not asked questions by a person who may have been bias in their own way or I may have answered differently based on what I thought may be the correct answer. A disadvantage to this method is that participants come from many varieties of life and different experiences that may lead them to differing paths. These researchers did a good job of incorporating this, however, there may still be some inconclusive studies based on many different factors.
To make this research even better, however, I think that these tests could have a small sample test before participants take the actual test. This would be for the participants to get accustomed to the way they are to answer the questions. This would eliminate inconclusive results where the participants get too many answers wrong for the data to be as helpful as it could be. This may give them a way to gain more conclusive results and more reliable answers.
Participant bias
Inductive approach
Survey
I decided to take the Presidential quiz because I was interested in what stereotypes I could have about presidents. I was actually stressed when I was getting ready to take it; there were so many directions that the site kept drilling into me and I started thinking that it might be more difficult than I originally thought, and I wanted to give correct answers. Then when I saw that I had to work as fast as possible, I tensed up some more – I’ve never enjoyed being timed or being rushed and the thought of being fast and accurate makes me worry about being too slow or not being accurate enough. Once the test actually started, though, I was beginning to enjoy it and it felt more like a game; when I made a mistake, I chuckled and shook my head and thought to myself that was a dumb mistake. Eventually when I had to identify good/bad along with Presidents Obama and Kennedy, I started to feel like I was Alex from A Clockwork Orange getting the Ludovico Treatment – being showed pictures of someone and words saying something bad and having to push a button for bad that is also a button for President Obama. But overall, I thought it was a good experience.
I wasn’t really surprised by my results of preferring Obama over JFK. Primarily because I’ve grown up with Obama and don’t really know what it was like living in JFK’s time. It was shocking to actually see it in black and white because I had never really thought about that kind of thing before. I don’t like the idea of comparing two people who have different places in history because I don’t know how either would handle the different situations that the other had faced.
The research methods that I noticed were survey, where I was asked personal preferences about certain things as well as general information about myself. I think that this experiment is a good example of random selection, because I had to show where I had heard about this survey and the surveyors would be able to see the wide variety of where people had come from.
Survey
Experiment
Random selection
I did not know what to expect as I opened up the webpage full of tests. I was a little overwhelmed so I just chose a random one and got going on it. Because I knew this was part of a study, I felt a lot of pressure taking this test. There was the whole sorting portion that made me nervous, because there were a couple instances that I got “errors” because my finger had a spasm and actually pushed the wrong key. I tried to not think about the fact that I was taking a test because I really didn’t want to affect my results at all. I wanted to get the most honest and accurate picture of my conscious or subconscious beliefs. I was pretty surprised by the results; I thought I was truly neutral on the subject but I ended up getting “slight” in one of the categories, and not in the direction I would have expected if I wasn’t as neutral as I thought. I ended up being in the 3rd-smallest category, at 10% of people reporting that result.
There weren’t a lot of variables in the actual test itself, unless you include the participant’s own actions. The test was multiple choice, so the responses were controlled. As far as the participant goes, however, they could have been distracted (i.e. listening to music or watching TV as they took the test), misunderstood one or multiple questions (because there is no one readily available to clarify any confusion) or simply didn’t take the test as seriously as they should have. There isn’t a way for the researchers to control this. I did my best to make sure I understood the questions before answering them (or in the case of the categorizing, I read the instructions every time, even though they were the same, just to be sure I didn’t miss anything) and tried to shut out distractions by taking the test in a quiet place with no one else around and no tv, music, etc. of any sort. The test itself seems like it’s very reliable; because it appears to be calculated by some sort of mathematical algorithm to tell you the outcome, there probably isn’t a whole lot of room for variation. Whether or not the test is valid or not is another question. This goes back to the inability to control the variables in and around the participant. If somebody messes up a lot because they are distracted, they likely won’t get an accurate result, and thus their result should not be used in the study as evidence one way or the other. If the researchers want more accurate and valid results, they should put the participants in a controlled environment in which they would take the test.
Terms: variables, control issues, reliability, validity
When I read this description for taking an “Implicit Association Test” I did not know at all what to expect. This in a way helped me be less biased as I took the tests, hopefully giving me more accurate results, however, I was greatly confused during several parts of the tests.
I started off by taking the Mental Health Treatments IAT. I first of all didn’t quite understand what this test was trying to establish, if I favored therapy or medication, or if I felt that one was more beneficial than the other. After reading the page that explained my results this issue was clarified, however going into the test I wasn’t quite sure what I was even trying to accomplish. The test itself was quite confusing for me personally. I have a very slow reaction time to begin with, and timed tests or activities always make me anxious, often making me produce worse results. This was not helpful when taking this test, as reaction time I’m assuming was a major determining factor in our implicit association. Figuring out what exactly was expected of me each round took me a couple of tries to get the hang of and by the time I completely understood what I was doing, the format changed and I was confused all over again. At least for me personally, I feel like this could have skewed my results. Several answers I chose could have simply been because I made a mistake as opposed to how I actually feel or would have responded, giving me a possibly different result. Nonetheless, I still feel this test was very insightful and provides more information into my implicit associations than I would have known otherwise.
In terms of my results, I relatively expected the results I received for the Mental Health Treatments IAT, that I tend to feel medications are more beneficial than therapy. However, when it came to the Alcohol IAT, I don’t necessarily agree with the results that that test produced. I have never drank in my life, and do not like to associate myself with drinking, alcohol or anything pertaining to that kind of environment. However, the results of this IAT suggested that I associate myself more with drinking than I do abstaining from drinking. This just seemed absurd. As I explained earlier, my results could have been skewed simply due to my confusion, and that could possibly explain this strange result.
Some research methods that I noticed when taking the test were reliability, variables, and participant bias. I have already kind of touched on reliability but I also want to elaborate on that it could have not been entirely reliable because the system did not even let you be “wrong.” It forced you to pick the seemingly correct answer, again another possible way for the results to be skewed and a false sense of causation to be present. In terms of variables, in each of the tests they seemed to be pretty obvious and controlled, especially the independent variables. Each test changed the independent variable each time to help produce results. For example, in the Mental Health Treatment IAT, they changed the independent variable from therapy to medication and assessed how I associated various words with these. Another aspect of research I noticed was possible participant bias. With each individual going in knowing they are going to be assessed on stereotypes and how they feel about these, this could cause them to subconsciously influence their answers to produce results more towards the stereotype they explicitly feel they lean towards.
Terms: reliability, causation, independent variable, participant bias
I have never taken an Implicit Association Test before so this experience was quite new to me. Before taking any of the tests, I thought about what the results would bring. I wondered if I would actually learn anything new about myself or if the results would simply reaffirm my own knowledge of myself. To try and determine this, I took three of the IATs. I picked weight as my first option because I know there is a huge concern about our country's overall health decline and I thought it would be interesting to see if my own perceived thoughts aligned with the test results. After that, I allowed the computer to choose for me. I got the gender-science test and the disability test.
Some of the experiences I had while taking the tests were the same for all three. I felt very frustrated when I was taking them. I think I was frustrated because I didn't want to get one wrong and I felt a lot of pressure to get all of them correct even though nobody was watching me take them, and I knew I wouldn't be graded on how many I got right. I also thought that changing the variables made it more difficult during the process. For example, they would switch which side each variable was on. They would also pair variables and then later switch the categories making it more challenging. I felt as though the test just confused me and frustrated me more than it found anything out about me because of this. I also think that the test may have found a correlation in my answers although I do not think they were caused by my thoughts but rather by my difficulty in remembering which side was which.
Most of the results produced by the IAT were in conjuncture with my own knowledge of myself, however the disability one was not. I was not very surprised with the results though. I thought they would be very generalized and reflect the opinions of the average population as a whole. I think that my guess was correct based on the results that were given. In my opinion, I think these tests did a better job of showing that I got going too fast and my instinct was to click the same one as I previously had. I don't think that the tests really showed any stereotypical behavior I may or may not have. I realize that they in fact could show that, however I would simply argue that it didn't.
The whole process did show signs of using research methods though. The most apparent thing to me was the use of experiments. All of the tests were experiments because they changed the independent variable in order to come to their conclusions. They also used surveys toward the end of the tests to get some information from the users. It is important for them to do that because they can then use the data they collected in order to find any possible correlations between the different answers. For example, maybe location affected a person's ability to recognize and side with a certain president. There are many biases which could affect the outcomes of the tests. I think that the tests took that into account because they asked questions about the opinion of the person taking the test and what they think the general public's opinion was.
Vocabulary Terms: variable, experiment, correlation, biases, survey
Since this was the first time that I would be taking tests that would be used for research purposes, I felt pressure on me to somehow do well. When I realized that the tests would tell me about what stereotypes I held about certain topics, I kind of didn’t want to take the tests anymore because I didn’t want to know what my results were. On the other hand I was interested in what the tests would say about me as person. While taking the results I felt apprehensive because I didn’t want to mess anything up and I was overthinking too much about the task itself. One of the two tests that I took made choose whether a word that was presented belonged under female or male groups. For example if the word was aunt it would go under the female group. As I continued I would have to do the same thing except for words that were either strong or delicate. I then had to answer a questionnaire about whether I associated females as delicate and males as strong and vice versa. My results suggested that I most often see males as strong and females as delicate. I wouldn’t agree completely on this result because I see females and males as both strong and delicate. I also don’t agree on the way they tested this because they would switch the male and females groups from side to side which made me focus more so I went slower even though I was supposed to go as fast as I could. At this point I was more worried about getting things right rather than being fast. My guess is that the dependent variable was the time it took me to complete each part of the test. The test went better as I went on and so I decided to take another one but I wasn’t able to choose either one of my tests.
While taking my tests I saw that one variable always changed whether it was the words or the sides the words were on. After both of the tests I took, I also had a survey to answer that asked questions pertaining to the test itself. I found the questionnaire to be very straightforward since it basically asked me whether I thought males were strong and females were delicate. I did my best not to have any bias but in tests where the participants know what the test is measuring then it is hard not to have any participant bias. I assume that the questionnaire was put in place to show correlations between the test itself and what the participant shared in the questionnaire.
Vocabulary: Correlation, dependent variable, variable, survey, participant bias.
For this experiment, I took the Family – Career Implicit Association Test (IAT). I was tasked to sort a series of words into the categories of either “Family” or “Career”, as well as to label several names as either “Male” or “Female”. The experiment was timed, and since I was required to respond “as quickly as I can”, I took it a little too literally and for 5 minutes I glued my eyes to the screen while pressing “E” and “I”. I was so focused on the task on the screen that when the categories switch “Male and Career” from left to right or shift “Male and Family” to the same column, it really did not affect me all that much. In fact, the rare times that my pace was slowed down were at the beginning when I felt that some words could be classified into both columns (E.g., “management” could mean “household management” or “business management”). I then thought back to the instructions and just went with my gut feelings since I realized I was overanalyzing. At the end, my result indicates “little or no association between Male and Female with Career and Family”, which places me in the median group with 17% of respondents.
I agree with my result, for several reasons. First of all, since I completed the task almost machine-like with little variation across sections, the result reflects this naturally. Secondly, as I was filling the questionnaire at the end of the experiment, I realize that I indeed have no association between genders and the family-career dichotomy. I studied overseas at a young age and thus was “liberated” early the many forms of family influences, including the prevalent ideas of men being the main source of wealth for the family. Therefore, there is no (conscious or unconscious) thought that influences my decisions during the task.
I recognized several aspects of the research methods from the experiment. In the IAT that I participated in, researchers maintain the categories (“Male”, “Female”, “Career”, “Family”) and their corresponding words and names as the independent variables. At the same time, they measure my responding time in different arrangements of categories as the dependent variables. I think that the experiment is reliable to a certain extent. If I were to do this IAT again with more or less the same level of concentration and external distraction, I am confident that I would be able to achieve the same results again. However, when taken control issues into consideration, I think it is relatively difficult to replicate the same experimenting condition. Perhaps I might be more fatigued the second time, or some disturbance could disrupt my focus on the task. All of this could impact my reaction time and consequently, my result.
Terms:• Variables (independent & dependent)
• Reliability
• Control issues
I took the Obama President IAT for this assignment. I had to sort pictures of Presidents Obama and Reagan into categories by pressing either e or i and then I had to sort words into good or bad categories, and then the two categories were combined. My results were that I showed no automatic preference between Obama and Reagan, which was the smallest percentage. Most other respondents preferred Obama to Reagan. I wasn't really surprised by my results. I respect both men and think they are doing the best they can to do the hardest job in our country.
The test was really hard though! I tried to go as fast as I could but I kept making mistakes. At one point, they switched around the categories and that was even harder. It was a good exercise for my brain, and I actually enjoyed it.
This test really interested me in how it explored people's biases. They did a good job with setting up a control group of pictures and words and then putting them together to see what I would associate with each president. I was part of an experiment, and it was really exciting for me to experience what we learn about in the classroom in the real world.
Terms: bias, control group, experiement