Chapter 4 Stage One: designing and developing an online Implicit Association Test to measure stereotypes of empathy in scientists
4.4. Step III: Creating and testing the SE-IAT website 1 Web versus Lab-based research
4.4.2. Creating the SE-IAT website
4.4.2.1. Brief overview of major existing software packages for timed test
There are a number of reaction time packages, both commercial as well as open-source, that are available for computerizing IAT nowadays. Either option comes with a certain trade-off. Open-source software packages usually require researchers to have knowledge of certain programming language(s) in order to create a customized test. Whilst most commercial software packages offer a user-friendly interface with drag- and-drop features, they do not require any programming skills. But those commercial software packages are often not offered online like most open-source packages. Table 4.5 summarised the advantages and disadvantages of current available reaction time software packages.
Table 4.5 A brief overview of reaction time software packages (Cek, 2015)
Softwares Online? Requiring
Programming skills?
Advantages Disadvantages
Commercial
DirectRT No Yes Simple syntax to follow to customize test Can only be used offline
E-Prime No No No programming skill needed Can only be used offline; need to
become acquainted with a multi-window visual user interface first
Inquist for Web Yes Yes Can be used with third party questionnaire
software such as Survey Monkey
Requiring relatively complex syntax to program tests
Superlab No No No programming skill needed;
questionnaires can be implemented at the same time
Can only be used offline
Open-source
FreeIAT No No Avoids Internet reaction time issues Can only be used offline; available only
with Windows system
WebIAT Yes Yes Tested on most systems and browsers Requiring researchers to host a Web
server
WEXTOR Yes No Fulfills all ethical requirements for web-
based studies
Hasn't been tested for reaction time studies
Softwares Online? Requiring
Programming skills?
Advantages Disadvantages
programming
jsPsych Yes Yes Fulfills all ethical requirements for web-
based studies
Requiring researchers to host a Web
server; requiring extensive
programming skills
Tatool Yes Yes Avoids Internet reaction time issues Requiring complex programming
knowledge; incompatible with third party questionnaire software
WebExp Yes Yes Fulfills all ethical requirements for web-
based studies
Requiring extensive programming
skills; incompatible with third party questionnaire software
JATOS Yes No Very simple user-friendly graphical
interface
Hasn't been tested for reaction time studies; incompatible with third party questionnaire software
ORTEngine Yes No Very simple user-friendly graphical
interface
Less flexible; requiring complex programming knowledge to be used with third party questionnaire software
ScriptingRT Yes Yes Validated for reaction time studies Requiring extensive programming
knowledge; incompatible with third party questionnaire software
4.4.2.2. Using the Project Implicit virtual lab to implement SE-IAT online
There are three requirements when selecting the appropriate software to create the SE- IAT website. First, due to the time constraints of the present PhD project, it is impractical to expect the researcher to put much effort in learning a new programming language from scratch, thus software packages requiring programming skills including
Inquisit for Web, OpenSesame, Webexp, ORTEngine, and ScriptRT were ruled out.
Second, the SE-IAT needs to be administered with self-report questionnaire, reaction time software packages that cannot be used with third party questionnaire software including Tatool and JATOS were not selected. Third, software packages that can only be used offline including DirectRT, E-Prime, Superlab and FreeIAT were dropped as well. The remaining open-source software packages WebIAT and WEXTOR fulfill all the three requirements, but neither was selected because WebIAT requires researchers to host a personal web server and WEXTOR had not been tested with any reaction time task before.
Under such circumstances, the task of creating a website to administer SE-IAT and self- report questionnaires online was outsourced to the technical service team from Project Implicit (PI), a virtual lab managed by IAT experts from Harvard University. The homepage link to PI is implicit.harvard.edu.
The Project Implicit Virtual Lab was selected because 1) both implicit measures and questionnaires could be easily implemented in one web browser; 2) the virtual lab possesses a mature and stable implicit measure infrastructure that has been used and tested successfully by many other researchers; 3) anonymity and security in transmission of individual responses are secured and 4) large data storage is provided.
Before implementing the study, the researcher was required to sign a contract with the PI team to set up the number of studies and the time length for data collection. In the current project, the researcher was allowed to conduct two rounds of data collection.
Before each round of data collection, the researcher would be given a test link to examine and modify all the implemented research materials. The website would be valid for only one year.
The first test link to the website was sent to the researcher on 14th June, 2016. The test link is as below:
https://app-prod-
03.implicit.harvard.edu/implicit/Launch?study=/user//nlofaro/contract.qin.empathy/m anager.expt.xml&refresh=true
Figure 4.1 is a screenshot for the incompatible pairing block of the online SE-IAT. When conducting the SE-IAT task, the category labels appear on the top left and right of the computer screen to remind participants of the response key mapping rules. The target and attribute labels are in different colours for the purpose of enhancing the distinctiveness of nominal features of given stimulus. Specifically, the Liberal arts and
Science labels and items are in green while the Rationality and Empathy labels and
items are in blue. The stimulus items for sorting are put in the middle of the screen. When a stimulus item is correctly categorised, a new item will immediately appear. But when the stimulus item is incorrectly categorised, a red "X" below the stimulus item will appear and the participants are obliged to fix the error before proceeding.
Figure 4.1 A screenshot for the SE-IAT incompatible pairing block