The Plan!

In essence, I want to build an online open-source platform (OAIG for Open Automated Item Generation) that will allow researchers to create tests using R or Python code for image creation related to their test items. Researchers will have access to open-source code (php and javascript) via GitHub that they can use on their own servers to administer their tests (and those from other researchers). The current vision is to use a matrix paradigm like the Raven’s where the number of rows and columns vary depending on the design - not unlike the KOTA which has two rows and four columns.

KOTA example:

Researchers also will have access to code repositories for image creation. BTW, images can contain text/numbers which allows for the use of verbal and math-based items. Depending on the application, items could be created on the fly or called from a file system.

SMAIG example:

Why? Researchers have few quality options when it comes to assessing cognitive abilities due to the proprietary nature of most tests. In addition, the concerns around test security and potential practice effects prevent open approaches to assessing psychological constructs. This type of approach allows and encourages people to practice the test. Also, I’m still testing the hypotheses that practice can help to reduce error and that some portion of said error is due to racial discrimnation.

Dream list: A team of PHP/Javascript programmers with SQL and GitHub expertise (R and/or Python experience a plus). A project manager with GitHub expertise!! At least two Ph.D.-level I/O psychologists with expertise in psychometrics.

Grant potential?

The Prototype

Progress has been slow this summer!! I initially thought I could use much of the code from the KOTA project but that isn’t going to happen. I’m now slogging through writing the fundamental structure of the system.

Did you find this page helpful? Consider sharing it 🙌