Hello to all.

I am in process of completing a “pilot study” to evaluate the utility of a novel “instrument cluster” for use in motor vehicles and other dynamic environments. The study design has two independent variables, including the novel instrument cluster and a “traditional” instrument cluster. The single dependent variable is participant response time.

A laptop with a 14” diagonal measure display is used to present stimuli to participants (see below). Participants provide responses to stimuli using the keyboard on the laptop. Software running on the laptop presents stimuli to participants and captures the speed and accuracy of their responses.

Instructions to participants include the request that they respond to stimuli (see below) as quickly as they can. They are also asked to respond accurately. This request is repeated twice during the instructions given to participants. (The rules for responding are also provided, and the participants are shown examples of the two different instrument clusters used in the task.)

The study simulates driving a vehicle. Participants see a road sign, followed two seconds later by an image of one of the instrument clusters. Participants are asked to respond “yes” if the combination of the road sign and the instrument cluster, or the instrument cluster alone, indicates they must “do something” as the driver of the vehicle. They are asked to respond “no” if the road sign and instrument cluster do not require action on their part.

A blank image appears on the laptop’s display for two seconds after the participant enters a response. Then the next road sign/instrument cluster pair is presented. This second pair uses the same instrument cluster as before, but with different “readings.” The second pair also includes a different road sign. This procedure for presenting stimuli and capturing responses continues until the participant has seen and responded to 16 different road sign/instrument cluster pairs. The participant is then allowed to take a short silent break. The process is then repeated, using the other instrument cluster in each pair.

Participants are assigned a “participant number” before they begin the study process. Participants who have an odd participant number see the novel instrument cluster in their first set of sixteen response pairs. Participants with an even participant number see the traditional instrument cluster in their first set of response pairs.

Each road sign is paired with instrument clusters that display the same readings. For example, when a “70 MPH” speed limit sign appears in either set, the novel instrument cluster and the traditional instrument cluster will show the same readings.

The correct response to eight of the item pairs is “yes.” Eight of the item pairs require a “no” response.

Item pairs appear in different orders for the two instrument clusters. These orders were developed using random number generation software.

It has been a number of years since I last needed to calculate statistics to evaluate the results of a study like this one. I believe this might be characterized as a “balanced” repeated measures design with, as above, two independent variables and a single dependent variable. This causes me to want to analyze the data using an appropriate ANOVA method, but I am not confident that that is the better method. I would appreciate input from the members of this learned community.

And then there is the other dependent variable I have not discussed – response accuracy. Obviously, there is an element of “shared” variance attached to this measure. That causes me to wonder if MANOVA is the better analysis to run.

Thank you for taking the time to read this. I will appreciate feedback that is offered.