For my money the Process Dissociation Procedure (PDP) from Jacoby is the best method for separating the two systems as it pits one against the other, thereby ensuring that they are dissociated as much as possible.

I don't do active research in the area but am familiar with the concepts of system 1 and system 2, although it seems tricky to try and differentiate which system is used by a participant entirely through online platforms.  

One suggestion may be to have self-reports of how they solved each task and then map participant responses to whether it was an analytical or heuristic approach used, which requires some justification for how its determined a given strategy is analytic or heuristic.  A participant may also use a combination of systems during tasks depending on the tasks.  

In terms of cognitive load, it would be expected that if very low cognitive load was experienced a heuristic decision making system was in process, however, they could also default to heuristics with high cognitive load too while attempting to complete a task analytically.

Dual Process Models, that have become central to a lot of social cognition research, include an automatic set of processes with heuristics (system 1) and deliberate processes from careful thought, reasoning, premises, and logic (system 2).  One way you can tell which system someone is using is to have them complete a task both with and without a cognitive load (e.g., distraction).  If performance stays the same, then they were using automatic processes to complete it and you can infer that they were using system 1.  But if an excess cognitive load reduces performance, then you can infer your participants were using effortful processing that was interrupted (system 2).  Probably the most direct way to estimate which system participants are using, especially online, is to measure time to complete tasks.  When people are using system 1 they reach conclusions faster than with system 2.  Hope this helps some, Simon! ~ Kevin

May be You will see different error distributions in Your task depending of the system used.

Hello,

We have been using a measure we've called "GeoCat" to distinguish system one vs. system two categorizations. System one, we feel, can be represented by similarity categorizations based on matches of attributes; system two, represented by similarity categorizations based on matches between relations of attributes. This is all made clearer in several publications by myself (Brakel) and colleagues measuring differential categorizations under conditions of subliminality, stress, untreated psychosis, and even in birds and young children. All of these publications can be downloaded from Research Gate. The dates are 2000, 2002, 2004, and 2005, for the ones in which I am the first author and then there is Blaisdell (2011?) for the bird study and Bazan (2013?) for the one with psychotic persons. Prof Bazan has also worked with students doing online versions, I believe. She too is available through Research Gate. (If not I can put you in touch with her.  My email is [email protected].)

Hope this is helpful

Hi!

I would also suggest the time constraint. If you give people a limited time to do the task they will more likely rely on system 1 thinking. As for measuring which system is active, this is tricky and I do not have a solution for you right now.

For my money the Process Dissociation Procedure (PDP) from Jacoby is the best method for separating the two systems as it pits one against the other, thereby ensuring that they are dissociated as much as possible.

Thanks for all the answers! They were really helpful, also to get deeper in the to topic and find new information and sources. However I guess I have to study each approach a little deeper to be sure what is the optimal way to measure the active system.

@Garret, the PDP is dealing with Recollection and Familiarity. Is it licit to map these to system 2 and system 1?

Maybe this is interesting for you...

Just saw this is a presentation, there is also an article in Psychological Science, published this month

Conference Paper Measuring Intuition: Unconscious Emotional Information Boost...

I am not a memory researcher, but since SYS1 and SYS2 are broad umbrella terms for a range of phenomena, I believe they map well onto these processes, e.g. implicit memory can be passively/unconsciously learned associations that can be elicited outside of controlled awareness (e.g. word priming), whereas explicit memory requires awareness and is prone to load etc. But difficult to say how well it would integrate into your task without more information.