Wextor Steps for a Simple Experiment
In this lab, we are going to create a simple experiment that people, anywhere in the world, can participate in simply by logging on to the website for your study. Remember that a simple experiment is not a survey. For example, an experiment must involve a manipulation: some participants will see one stimulus, other participants will see a different stimulus. Your manipulation can consist of different instructions, different wordings of problems, different words, or different images.
1. Go to http://wextor.org/wextor/en/
2. Once there, under “need an account?” will be a “sign up” button.
Click on that button. It should take you to http://wextor.org/wextor/en/register.php
Follow the instructions to register at the site. You will need to make up a login name and a password, you will need to provide an e-mail address, and you will need to accept the “Terms of use.”
3. Go to the FAQ section (http://wextor.org/wextor/en/faq.php). Go through the tutorial (http://wextor.org/wextor_docs/WEXTOR_tutorial.pdf). That tutorial will help you understand factorial designs.
Then, look at the techniques and tricks page.
4. Log in to Wextor (you will need the activation code that was e-mailed to you after you registered) and start working on your own simple experiment by clicking on the first bulleted link:
· Create/modify an experimental design
Start by putting the name of your experiment in the “Name Your experimental design” box. Then, click the “create” button. Then, go through the 9 steps. The handout below will help you, but don’t be afraid to ask your professor for help.
Step 1 “Defining factors”
How many between-subjects factors does your experimental design consist of?
Because you are doing a simple experiment, you have only 1 independent variable.
How many within-subjects factors does your experimental design consist of?
NOTE: Currently, it is best to define only one within factor. If you have multiple within factors in your design, then please redefine their combinations of levels as levels of one within factor, e.g. color (red, blue) and font size (10, 12) of Web pages as appearance (red-10, red-12, blue-10, blue-12).
Because you are doing a simple experiment, you no within-subjects factors (your only independent variable is between subjects), so you would put zero.
Does your experiment include at least one quasi-experimental (natural) factor?
For a simple experiment, you would select “no.” The simple experiment has only one predictor—and that factor is manipulated by the experimenter.
Click on the > button.
Your experiment consists of 1 factors. Please name all factors and indicate the number of levels for each factor.
Please name your between-subjects factors and indicate the number of levels for each factor.
Number of levels: 2
Name: Independent variable
Note that for a simple experiment, you will have two levels.
For a multiple-group experiment (see Chapter 11), you will have at least 3 levels.
For a factorial experiment (see Chapter 12), it will depend on what type of factorial you have. For a 2 X 2 factorial, you would have 2 levels of each factor; for a 3 X 2, you would have 3 levels of the first factor and 2 levels of the 2nd factor.
Please name the levels of the factors you defined in Step 2.
Please name the levels of your between-subjects factors!
Name:
Name:
Please name your experimental conditions. We already made a naming proposal, which consists of the abbreviated form for the experimental condition and four random characters. Please note that it is wise to NOT use names that reveal information about the underlying structure of your Web experiment.
If your experimental design is incomplete, you may now erase the superfluous experimental conditions. Simply uncheck all experimental conditions you don't need.
Experimental condition 1:
Experimental condition 2:
How many pages does one experimental condition consist of? If you are intending to conduct this experiment on the World Wide Web, consider using the warm-up technique for drop-out control. Please note that the number of Web pages has to be the same in every experimental condition - otherwise we would introduce a confounding variable.
WEXTOR automatically generates five Web pages:
index page: this is the homepage of your experiment.
start page: we need to test whether a participant's Web browser is JavaScript compatible. If the test result is positive then the user will be redirected to the start page automatically. This page should contain general information about your Web experiment. If your instruction requires a participant consent form, then this page would be the place for it.
source page: If the user doesn't have a JavaScript capable Web browser, he will be sent to the source page. This page should display the information that the user won't be able to participate in this experiment because the Web browser isn't JavaScript capable. Further you can tell the user where to get a qualifying Web browser or how to enable JavaScript in the browser preferences.
demos page: this is the place to ask the user for demographical information.
thank page: this is the place to thank participants for their participation.
NOTE: Do not include the 5 default pages with the number of pages you fill in.
Number of Web pages:
Time-out in seconds.
(0: the page will not time-out)
Please name your Web pages now (e.g., intro.html).
Name of index page:
Name of start page:
Name of source page:
Name of demos page:
Name of thank page:
Time out in seconds= 0
Name of Web page 1:
Choosing obvious file names is a frequent error when creating and designing Web experiments. For this reason WEXTOR offers to add some random characters at the end of your file names. Simply click on the button 'add'. You will see the changes immediately in your file names.
Your experimental design does not include a within-subjects factor. You can skip this step!
Your experimental design does not include a quasi-experimental (natural) factor. You can skip this step!
Your experimental design is finished now. You have the possibility to take a look at your experimental design, and the visual display thereof.
When you click the download link, your browser downloads the zip file to your computer. Be sure to note where the file is saved. The file is compressed and archived. If it doesn't decompress and unarchive automatically, you must expand it using a program such as StuffIt Expander (Mac, Windows, and Linux) or WinZip (Windows). If you don't have StuffIt Expander, you can get a copy by visiting http://www.stuffit.com/downloads.html. There is no charge for StuffIt Expander.
· list of your experimental design
· visual display of the experimental design
Choose a CSS style sheet for your experiment:
If your computer doesn’t automatically uncompress those files for you (it did, when I tested it), go to www.wobzip.org to uncompress your files.
5. One problem with Wextor is that it does not automatically you’re your two conditions different. Therefore, you will need to modify one of your files using “Text Edit” or “Notepad” so that your “two” conditions are different. So, you will probably need to double-click on the hard drive icon, double-click on the "Applications" folder, find the Text Edit program, double-click on it, and then open your "experiment" file of your second condition with Text Edit
6. After you have made sure your experiment does what you want, zip your files. You can probably do this by simply clicking on the folder and then, under the "File" menu, selecting "Create an archive for ...." If that does not work, either of the following web sites will zip your files for you.
7. Once you have a zipped file (the file's name will end with ".zip,"go back to Wextor, and choose the second bulleted link:
· Upload an experiment, download data.
After you click that link, Wextor will tell you how to upload your zipped file.
Wextor will confirm that your file was successfully loaded and will tell you the URL for your experiment.
8. E-mail that URL (e.g.,
http://wextor.org:8080/dlogue/loguexperiment/)to me.