Marsh, R.L., Hicks, J.L., Cook, G.I., Hansen, J.S. & Pallos, A.L. (2003). Interference to ongoing activities covaries with the characteristics of an event-based intention. Journal of Experimental Psychology: Learning, Memory,& Cognition,29, 861-870.
In a nutshell prospective memory is the type of memory involved with the intention that we have to complete tasks, large or small, in our lives. There are several types of prospective memory.
· Time Based – I need to set my VCR to tape Las Vegas at 8 o’clock.
· Activity Based – As soon as I finish my homework I will call Frank the Tank.
· Novel Task – I will go to the Pearl Jam concert in July.
· Habitual intention – I have to take my pill tonight.
· Event Based (EBPM) – I need to call the cops if I see Charlie Manson on the street.
For the purposes of this study the authors dealt with EBPM, as have many previous studies. However, the novel angle this time is the examination of interference to the ongoing task. Most of the previous EBPM studies looked at variables that affect the cue detection. This study looks more at the variables that affect the latencies in the ongoing task.
Prospective memory can be broken down into four components: recognition of the prospective cues, verification of said cue, retrieval of the proper response, and coordination of this prospective process and the ongoing process. The verification and retrieval processes are often thought of as the retrospective parts of the prospective whole. The goal of the current study was to see how these retrospective components affect the ongoing task.
Experiment One
Method
A simple lexical decision task was used for the ongoing task in all four experiments. Words and pronounceable nonwords were presented to the Ps, with specific key presses assigned to each. In the single cue condition the word dog replaced words in the list eight times as the prospective cue. In the categorical-intention condition eight animal words (e.g. monkey) were presented in place of eight of the list words. In both prospective conditions, the participants were asked to respond to the prospective cues by pressing the “/” key after responding to the word/nonword task. Upon the completion of instructions the Ps were given a five minute distracter before the ongoing task began to prevent any vigilance element to interfere. This distracter task was administered in all experiments.
· Performance was better in the single cue condition than in the category condition.
· The RT for the classification of nonwords was slower, suggesting that the Ps were taking the ongoing task seriously.
· The categorical prospective task slowed response to the ongoing task significantly over the control, while the single cue condition did not.
· This increase in monitoring in the category condition did not result in higher prospective success, meaning that attentional allocation does not guarantee success.
· Examination of the difference scores suggests that there is no difference in the interference caused by the different prospective conditions. In turn, this suggests that the monitoring effect and the interference effect may be independent.
Experiment Two
Method
This experiment was used to manipulate the size of the cue-set. Ps learned sets of four or eight prospective cue words to criterion. As a second variable check the lists were either semantically related or not semantically related. Marsh et al. point out the important fact that the words were matched as closely as possible on the different list on typical word characteristics, such as word frequency, syllable length, etc.
· No main effect of cue-set size in the accuracy scores.
· There was a main effect of cue-set size and cue relatedness in the latency measures, with the large sets and the unrelated sets causing longer latencies.
· The difference score showed higher interference in the ongoing task caused by the unrelated cue-sets, while the set size did not. This suggests that the cue-set size did not affect the verification process.
· The overall interference in this experiment was higher, which makes sense in light of the differences in the cues used.
Experiment Three
Method
This experiment was used to test for any interference that may arise during target retrieval. In order to test for this interference three conditions (in addition to our old friend “control”) were used. In the aptly named cue-target conditions Ps learned cue-target word pairs to criterion. In the first of these conditions the cues and targets were semantically related, while the cues and targets were (you guessed it!) unrelated in the second condition. The final condition was a cues-only condition. The response required in this experiment was a verbalization of either the target associates (cue-target conditions) or the cue itself (cues-only condition).
· The latencies in the ongoing task for each of the three prospective conditions were significantly slower than the control condition.
· The related cue-target pair was faster than unrelated pairs, with only marginal significance.
Experiment Four
Method
In the first three experiments the Ps were required to give their response to the prospective task only after giving their response to the ongoing task. So perhaps the retrospective processes could have taken place after the response to the ongoing task. Thus, it may be that some EBPM is automatic. Strong associations may be retrieved more automatically. To test this idea four conditions were used, two experimental and two control. Subjects in all conditions learned the same set of eight prospective cues to criterion. In the first control condition the Ps were required to respond by saying the word “now” whenever they detected a prospective cue. The second control condition required them to respond with whatever first associate came to mind. In the associated-pairs experimental condition Ps learned specific target associates for the prospective cues. These targets were their required verbal responses upon detection of prospective cues. The second experimental condition was the same except that the associates were switched so that they were not related.
· Prospective performance was worse in the unassociated condition than the associated condition.
· Interference was greater in the first associate control condition than the “now” control condition, which may be due to the novel retrieval required at each cue in the first associate condition.
· The first associate condition showed a latency that fell between the associated and unassociated conditions. This suggests that priming at study in the associated condition may have caused an automization of retrieval.
This article took on the impact of variables that may affect the latency of the ongoing task, as opposed to the variables affecting the success rate of prospective memory. Overall this study found (as did Smith 2000,2001) that the EBPM variables can affect the attentional allocation of individuals, which, in turn, affect the speed on the ongoing task. It is important to note, however, that while longer latencies may suggest greater prospective monitoring, the longer latencies do not guarantee higher prospective memory success rates.
Another important consideration is that fact that there was interference caused in the first place. If all prospective memory tasks were automatic and obligatory then no interference would occur with the ongoing task. The finding of interference, then, suggests that there may be multiple processes going on.
As a final statement and suggestion Marsh et al. mention that the relationship between the EBPM and ongoing task is reciprocal, with the nature of the ongoing task affecting the prospective memory task and the prospective memory task affecting response to the ongoing task.