Automatic effects of instructions do not require the intention to execute these instructions

TLDR: This article showed that maintaining instructed stimulus-response mappings for future recognition, rather than for future execution, can also lead to an instruction-based congruency effect, even when it is very unlikely that participants form the intention to execute instructions.

Abstract: Prior research established that newly instructed stimulus-response mappings, which have never been executed overtly before, can lead to automatic response-congruency effects. Such instruction-based congruency effects have been taken as evidence for the hypothesis that the intention to execute stimulus-response mappings results into functional associations that serve future execution. The present study challenges this hypothesis by demonstrating in a series of four experiments that maintaining instructed stimulus-response mappings for future recognition, rather than for future execution, can also lead to an instruction-based congruency effect. These findings indicate that the instruction-based congruency effect emerges even when it is very unlikely that participants form the intention to execute instructions. Alternative interpretations of the instruction-based congruency effect are discussed.

Full text

Running head: AUTOMATIC EFFECTS OF INSTRUCTIONS 1
Automatic effects of instructions do not require the intention to execute these instructions.
Baptist Liefooghe & Jan De Houwer
Department of Experimental-Clinical & Health Psychology, Ghent University
Author Note
This research was supported by grant BOF16/MET_V/002 of Ghent University to Jan
De Houwer.
Correspondence concerning this article should be addressed to Baptist Liefooghe, H.
Dunantlaan 2, B-9000, Ghent, Belgium. E-mail: baptist.liefooghe@ugent.be

AUTOMATIC EFFECTS OF INSTRUCTIONS 2
Abstract
Prior research established that newly instructed stimulus-response mappings, which have
never been executed overtly before, can lead to automatic response-congruency effects. Such
instruction-based congruency effects have been taken as evidence for the hypothesis that the
intention to execute stimulus-response mappings results into functional associations that
serve future execution. The present study challenges this hypothesis by demonstrating in a
series of four experiments that maintaining instructed stimulus-response mappings for future
recognition, rather than for future execution, can also lead to an instruction-based
congruency effect. These findings indicate that the instruction-based congruency effect
emerges even when it is very unlikely that participants form the intention to execute
instructions. Alternative interpretations of the instruction-based congruency effect are
discussed.

AUTOMATIC EFFECTS OF INSTRUCTIONS 3
Automatic effects of instructions do not require the intention to execute these instructions.
In recent years a growing body of research focused on the assimilation of new
instructions and how this leads to action. The currently prevailing view is that instructions,
intended to be executed, can be implemented into a procedural representation in working
memory, which guides their execution by enabling reflexive behavior (e.g., Brass, Liefooghe,
Braem, & De Houwer, in press; Cohen-Kdoshay & Meiran, 2007; Liefooghe, Wenke, & De
Houwer, 2012; Meiran, Cole, & Braver, 2012; Wenke, Gaschler, & Nattkemper, 2007). In
analogy to Exner’s (1879) notion of the “prepared reflex”, the implementation of instructions
thus leads to a state of preparedness, which Meiran et al. (2012) labeled intention-based
reflexivity.
The study of intention-based reflexivity mainly uses procedures in which the automatic
effect of new and merely instructed Stimulus-Response (S-R) mappings is measured (e.g.,
Braem et al., 2017; Cohen-Kdoshay & Meiran, 2007, 2009; Liefooghe et al., 2012, 2013, 2016;
Meiran et al., 2015a, 2015b; Theeuwes et al., 2014; Wenke et al., 2007, 2009, 2015). For
instance, Liefooghe et al. (2012) presented participants with different runs of trials on which
two tasks had to be performed that shared stimuli and responses: the inducer and the
diagnostic task. At the start of each run, participants received two novel arbitrary S-R
mappings of the inducer task, each assigning a stimulus either to a left or a right response
based on the identity of the stimulus (e.g., If “X”, press left; if “Y”, press right). Before
executing the inducer task, several trials of the diagnostic task were performed, on which
participants decided whether a stimulus was presented in italic or upright, again by pressing
a left or right response key (e.g., upright, press left; italic, press right). After a number of
trials of the diagnostic task, a probe stimulus of the inducer task was presented. Liefooghe et
al. (2012) observed that performance in the diagnostic task, in terms of speed and
sometimes in terms of accuracy, was better when the correct response on the diagnostic task
matched with the instructions of the inducer task (e.g., “X” presented upright or “Y”
presented in italic) than when the correct response on the diagnostic task did not match

AUTOMATIC EFFECTS OF INSTRUCTIONS 4
with the S-R mappings of the inducer task (e.g., “Y” presented upright or “X” presented in
italic). Given that (1) the diagnostic task was performed immediately after the presentation
of the instructions of the inducer task, thus prior to the overt execution of these instructions
and (2) the inducer task comprised novel S-R mappings on each run, the conclusion was
drawn that the congruency effect observed in the diagnostic task was based on the instructed
S-R mappings of the inducer task, which were never executed overtly before.
Several findings suggest that instruction-based congruency effects offer a genuine proxy
of intention-based reflexivity. Wenke et al. (2009) observed that the instruction-based
congruency effect disappeared when participants were frequently signaled that the inducer
task would not proceed, which suggests that participants refrained to implement the
instructions under such conditions. Liefooghe et al. (2012) observed an instruction-based
congruency effect when the inducer task required the execution of the instructed S-R
mappings (see supra), but not when the inducer task required the mere recall or recognition
of the instructed S-R mappings. Liefooghe et al. (2012) proposed that when instructed S-R
mappings are maintained for future recall or recognition (i.e., when no execution intention is
present), they remain in a declarative format in working memory, which does not induce an
instruction-based congruency effect. In contrast, when participants have the prospective
intention to execute the instructed S-R mappings, a procedural representation of the
instructed S-R mappings is formed, which induces an instruction-based congruency effect.
This conclusion was furthermore supported by the observation that distinct brain regions are
recruited during the implementation as compared to the mere memorization of instructions
(Demanet et al., 2016), as well as, by research demonstrating that the instruction-based
congruency effect is underlain by pre-motor activation (Everaert et al., 2014; Meiran et al.,
2015b). Finally, several studies observed that the instruction-based congruency effect is a
function of the degree by which the inducer task is prepared for (Deltomme, Liefooghe, &
Braem, submitted; Liefooghe, De Houwer, & Wenke, 2013; Meiran et al., 2015a, Experiment
4), which also suggests its dependency on the intention to execute instructions.

AUTOMATIC EFFECTS OF INSTRUCTIONS 5
The observation that the instruction-based congruency effect is a function of intention
and preparation, furthermore, contrasts with findings concerning the task-rule congruency
effect, based on overtly practiced S-R mappings. The task-rule congruency effect is observed
when participants have to switch between two overlapping tasks (i.e., task switching; Kiesel
et al., 2010; Monsell, 2003; Vandierendonck, Liefooghe, & Verbruggen, 2010). As for the
instruction-based congruency effect, the task-rule congruency effect denotes the difference
between congruent trials on which both tasks point toward the same response and
incongruent trials on which both tasks point toward a different response. Yamaguchi &
Proctor (2011) observed a task-rule congruency effect when both tasks were presented in
separate blocks of trials, which suggests that this effect is present even if the intention to
execute the irrelevant S-R mappings is minimal. In addition, the task-rule congruency effect
has been reported to be independent of the degree by which an upcoming task is prepared
for (e.g., Meiran, 1996).
Whereas most studies converge towards the hypothesis that the instruction-based
congruency effect reflects intention-based reflexivity, some findings, however, suggest that the
instruction-based congruency effect may also be induced when the intention to execute
instructions is absent. In contrast to what was reported by Liefooghe et al. (2012), these
findings show that an instruction-based congruency effect can be obtained when instructions
are merely maintained for future recall. More specifically, Theeuwes, De Houwer, Eder, and
Liefooghe (2015) investigated whether an instruction-based congruency effect can also be
obtained on the basis of contingencies between a response and the effect this response
produces in the environment (i.e., an action effect). To this end, the procedure of Liefooghe
et al. (2012) was adapted, such that each run started with the instruction of two new
Response-Effect contingencies (e.g., if you press the left key, the letter Q will appear on the
screen; if you press the right key, the letter P will appear on the screen). In the diagnostic
task, the effect-stimulus letters of the inducer instructions (i.e., “Q”,“P”) were used as
targets, the orientation of which had to be judged by pressing a left or a right key. An

Frequently asked questions

Q1. What are the contributions mentioned in the paper "Automatic effects of instructions do not require the intention to execute these instructions" ?

The present study challenges this hypothesis by demonstrating in a series of four experiments that maintaining instructed stimulus-response mappings for future recognition, rather than for future execution, can also lead to an instruction-based congruency effect. Alternative interpretations of the instruction-based congruency effect are discussed. The study of intention-based reflexivity mainly uses procedures in which the automatic effect of new and merely instructed Stimulus-Response ( S-R ) mappings is measured ( e. g., Braem et al., 2017 ; Cohen-Kdoshay & Meiran, 2007, 2009 ; Liefooghe et al., 2012, 2013, 2016 ; Meiran et al., 2015a, 2015b ; Theeuwes et al., 2014 ; Wenke et al., 2007, 2009, 2015 ). This conclusion was furthermore supported by the observation that distinct brain regions are recruited during the implementation as compared to the mere memorization of instructions ( Demanet et al., 2016 ), as well as, by research demonstrating that the instruction-based congruency effect is underlain by pre-motor activation ( Everaert et al., 2014 ; Meiran et al., 2015b ). In addition, the task-rule congruency effect has been reported to be independent of the degree by which an upcoming task is prepared for ( e. g., Meiran, 1996 ). This is, for instance, illustrated by research, which focuses on the representation of serial order in working memory ( see Abrahamse, van Dijck, Majerus, & Fias, 2014 for a review ). A common assumption in this domain is that serial information is maintained by creating a spatial representation in working memory in which serial information is ordered from left to right. Evidence for this hypothesis was reported by van AUTOMATIC EFFECTS OF INSTRUCTIONS 7 Dijck and Fias ( 2011 ), who presented participants with words that appeared serially on the screen. Taken together, the aformentioned findings challenge the conclusions of Liefooghe et al. ( 2012 ), who claimed that the instruction-based congruency effect is absent when the inducer task only requires the memorization of instructions for future recall. To this end, the authors focused on one of the memorization conditions reported by Liefooghe et al. ( 2012 ), namely the visual-recognition condition. The present study, however, questions this null finding by considering two concerns about the study of Liefooghe et al. This issue is considered AUTOMATIC EFFECTS OF INSTRUCTIONS 8 in Experiment 1 of the present study. This issue was dealt with by changing the parameters of the inducer task across Experiments 1 to 4 of the present study. Several findings suggest that instruction-based congruency effects offer a genuine proxy of intention-based reflexivity. Wenke et al. ( 2009 ) observed that the instruction-based congruency effect disappeared when participants were frequently signaled that the inducer task would not proceed, which suggests that participants refrained to implement the instructions under such conditions. Finally, several studies observed that the instruction-based congruency effect is a function of the degree by which the inducer task is prepared for ( Deltomme, Liefooghe, & Braem, submitted ; Liefooghe, De Houwer, & Wenke, 2013 ; Meiran et al., 2015a, Experiment 4 ), which also suggests its dependency on the intention to execute instructions. The observation that the instruction-based congruency effect is a function of intention and preparation, furthermore, contrasts with findings concerning the task-rule congruency effect, based on overtly practiced S-R mappings. Yamaguchi & Proctor ( 2011 ) observed a task-rule congruency effect when both tasks were presented in separate blocks of trials, which suggests that this effect is present even if the intention to execute the irrelevant S-R mappings is minimal. Whereas most studies converge towards the hypothesis that the instruction-based congruency effect reflects intention-based reflexivity, some findings, however, suggest that the instruction-based congruency effect may also be induced when the intention to execute instructions is absent. The results of Theeuwes et al. ( 2015 ) indicate that an instruction-based congruency effect can be obtained when the inducer task simply requires the maintenance of instructions, which suggests that the intention to execute instructions is unnecessary to obtain an instruction-based congruency effect. The findings of Theeuwes et al. ( 2015 ) are, furthermore, in line with other demonstrations, which suggest that simply maintaining information in working memory is sufficient to elicit automatic response effects. Accordingly, the authors reevaluated the conclusions of Liefooghe et al. ( 2012 ) by further testing the extent to which the memorization of S-R mappings allows for an instruction-based congruency effect to emerge. 

Q2. What have the authors stated for future works in "Automatic effects of instructions do not require the intention to execute these instructions" ?

To this end, the authors reevaluated the conclusion that the instruction-based congruency effect can not be obtained when participants maintain instructed S-R mappings for future recall or recognition ( Liefooghe et al., 2012 ). The intention to actively maintain instructed S-R mappings for future recall or recognition may, however, also constitute a cross-talk inducing demand. Future research will be needed to pinpoint its exact nature. Their results suggest that the instruction-based congruency effect needs to be interpreted without calling upon intention-based reflexivity. 

Q3. How long did the S-R mappings remain on the screen?

750ms after the last response in the diagnostic task, two S-R mappings appeared, which remained on-screen for 5000ms or until participants responded. 

Q4. What is the effect of imposing a strict response deadline in the inducer task?

For instance, Liefooghe et al. (2013) assumed that imposing a strict response deadline in the inducer task would encourage participants to prepare more thoroughly for the inducer task and thus require them to maintain the instructed S-R mappings more actively during the diagnostic task. 

Q5. How many participants were excluded on the basis of very low accuracies?

Six participants were excluded on the basis of very low accuracies either in thediagnostic task (.61) or in the inducer task (.54, .55, .56, .54, .60). 

Q6. What is the effect of the instruction-based congruency effect?

The larger instruction-based congruency effect observed in Experiment 4 may suggest that when participants are encouraged to maintain both instructed S-R mappings and to prepare for the inducer task more thoroughly, the rate by which the instructed S-R mappings are refreshed during the diagnostic task is increased. 

Q7. What was the evidence for the instruction-based congruency effect?

When combining RT and PE by using LISAS, anecdotal evidence in favor of the presence of an instruction-based congruency effect was obtained. 

Q8. What is the main conclusion of the present study?

Taken together, the core conclusion of the present study is that the instruction-basedcongruency effect is not confined to the intention to execute instructions and the processes underlying this effect seem more complex than the mere conception of intention-based reflexivity. 

Q9. How many participants had extremely low accuracies on the diagnostic task?

Visual inspection of the data revealed that four participants had extremely low accuracies either on the diagnostic task (.50, .50, .51) or on the inducer task (.46). 

Q10. What is the alternative interpretation of the current findings?

An alternative interpretation of the current findings is that the intention to recall orrecognize instructed S-R mappings also leads to the formation of procedural representations, as it is the case for the intention to execute the instructed S-R mappings. 

Q11. What was the effect of the inducer task?

As was already mentioned in the Introduction, an instruction-based congruency effect was observed in this study, even though the inducer task was a recall task. 

Q12. What is the main message of the present study?

For now, the main message of the present study is that the authors need to be cautious when drawing conclusions about intention-based reflexivity on the basis of the instruction-based congruency effect. 

Q13. Why did the inducer task have a stricter response deadline?

Because Experiment 2 used a three-choice inducer task with a lenient response deadline, this experiment now served as the baseline for the between-experiment comparison. 

Q14. What is the evidence in favour of the hypothesis that an instruction-based congruency effect can?

the evidence obtained in favour of the hypothesis that an instruction-based congruency effect can be obtained when the inducer task requires visual recognition, was generally weak. 

Q15. What is the evidence in favor of the hypothesis that when the inducer task requires visual recall,?

when the inducer task requires visual recall, memorizing only one S-R mapping is not sufficient to obtain an instruction-based congruency effect.