3. Block duration

-Longer trials are difficult mentally for subjects

-The linearity assumption of the General Linear Model (a common fNIRS analysis method) is violated when saturation takes place. As such, blocks over 60s are not recommended. 

4. Addressing subject anticipation 

-With the same taskrepeatedly, the subject maylearn to anticipate the stimuli, resulting in reduced hemodynamic activation –e.g. a go/no go task doesn’t lend itself well to a block design study.

Event-Related Designs

In an event related design, each task is presented individually for a short amount of time –e.g. 3s. In this way, tasks can be more  randomized,  rather  than  being  blocked  together  by  condition.

Like block design, event related design also relies on the assumption that the hemodynamic response adds linearly – but the analysis is more complex, as the response to two or more different stimuli can occur simultaneously, and must be able to be  distinguished from one another during analysis (as inthe general linear model).

Considerations for Event-Related Designs

1.Time between events

-For an event-related design to be effective the time between two successive stimuli should be at least 2 seconds (not faster than 0.5 Hz).

2.Data analysis

-Analysis of the data is significantly more complex and dependent on accurate modelling of the HRF.

-A jittered duration (varied) rest period can increase the statistical power –e.g. 2s, 4s, 6s