CC BY
194
Вестник СПбГУ. Сер. 16. 2014. Вып. 4
ТЕЗИСЫ ДОКЛАДОВ КОНФЕРЕНЦИИ (на английском языке)
UDC 159.95
E. Norman, Ryan B. Scott, E. Jones, M. C. Price, Z. Dienes STRATEGIC CONTROL IN ARTIFICIAL GRAMMAR LEARNING
Elisabeth Norman — Associate Professor at University of Bergen, Bergen, Norway, Elisabeth.Norman@ psysp.uib.no
Ryan Scott — Lecturer at University of Sussex, Falmer, United Kingdom, r.b.scott@sussex.ac.uk
Emma Jones — Ph.D. candidate at Faculty of Psychology, University of Bergen, Norway, Emma.Jones@ psykp.uib.no
Mark Price — Associate Professor at Faculty of Psychology, University of Bergen, Norway, Mark.Price@ psysp.uib.no
Zoltan Dienes — Professor of Psychology School of Psychology University of Sussex, Falmer, United Kingdom, dienes@sussex.ac.uk
Traditionally, strategic control over the application of knowledge is taken to indicate that the knowledge in question is consciously accessible (Jacoby, 1991). In the context of implicit learning, it would indicate that learning is explicit rather than implicit. However, the assumption that strategic control is limited to consciousness has been challenged by recent empirical findings showing that unconsciously perceived stimuli can interfere with tasks traditionally thought to require strategic control (Lau & Passingham, 2007; van Gaal, Ridderinkhof, Scholte & Lamme, 2010). We address the question of whether unconscious structural knowledge of two grammars can be strategically applied. In an artificial grammar learning experiment all participants (N=120) were trained on two different grammars. Instructions as to which grammar to apply varied randomly between individual trials of a subsequent test phase (Norman, Price & Jones, 2011). The nature of each grammar was disguised by random variation in irrelevant properties of individual string elements. After each classification response, participants made a combined judgement: they had to report (1) whether their decision was based on a random choice, a feeling of intuition/familiarity, or an explicit memory/rule, and (2) which stimulus property (letters, colours, fonts) they thought or guessed were relevant to the grammaticality of the string. The main question of interest was whether participants could strategically control their grammar knowledge on trials that were based on implicit decision strategies and attributed to irrelevant stimulus properties. Results were reported at the conference.
UDC 159.95
N. Moroshkina
DECISION-MAKING IN CASE OF CONFLICT BETWEEN IMPLICIT AND EXPLICIT KNOWLEDGE
Nadezhda Moroshkina — Assistant Professor at St. Petersburg State University, St. Petersburg, Russia, moroshkina.n@gmail.com
The work is dedicated to the investigation of the interaction between subject's implicit and explicit knowledge acquired during learning. In the first phase of the experiment the conditions were established in which explicit criteria for stimuli classification were accompanied by changes in irrelevant parameters (hidden hint) that allowed implicit learning to occur. In the second phase presentation of the new stimuli also included the hidden hint, but in the quarter (group 1) or half (group 2) of cases this hint contradicted the explicit classification criteria. This critical situation helped to evaluate which of the systems — explicit or implicit — took a priority. Two experiments were conducted. In the first experiment subjects had to classify visually presented banknotes as counterfeit bills or real ones. The early noticeable shift of the banknotes from the center of the screen served as the hidden hint. In the second experiment subjects were presented with letter strings and were ask to guess without solving if those strings were anagrams or not. The tiny font changes served as a hidden hint. The following results were obtained. 1. In the first phase the hint contributed to learning, still remaining unconscious to subjects. 2. Only in case where 75% of new stimuli were accompanied by congruent hint and 25% — by conflict one (group 1) both positive and negative transfer of implicit knowledge to the new stimuli was observed. 3. If 75% of new stimuli were accompanied by conflict hint and only 25% by the congruent one, the transfer did not occur. We consider this effect to be the loss of trust in the hidden hint. In the course of task performance subjects started rely less on the intuitive decisions and guessing strategy and were trying to use explicit criteria.
Supported by SPbSU (grant 0.38.518.2013).
UDC 159.95
A. Poplawska, R. Sterczynski, M.Roczniewska
FOCUSING ON GOAL RELEVANCE — IS IT CRUCIAL FOR ARTIFICIAL GRAMMAR LEARNING?
Agnieszka Poplawska — Assistant Professor at University of Social Sciences and Humanities, Sopot, Poland, agnieszkapoplawska1@wp.pl
Radoslaw Sterczynski — Assistant Professor at The Warsaw School of Social Sciences and Humanities, Sopot, Poland, rsterczynski@swps.edu.pl
Marta Roczniewska — Assistant at The Warsaw School of Social Sciences and Humanities, Sopot, Poland, mroczniewska@swps.edu.pl
Eitam, Schul and Hassin (2009) reported studies in which focusing on goal relevance was a necessary condition for artificial grammar learning task. This investigation used a paradigm in which participants were exposed to sequences of stimuli that varied in two dimensions (colours and letters) i. e. letters were shown on colored background. The order
of background colour and the order of letters were determined by a different grammar. The special instruction was used for two groups of participants. In one group the attention was oriented to learning letters. In second group the attention was oriented to learning colours. In both groups the artificial grammar learning process was observed.
We used very similar paradigm in which we presented letters covered by different colours. We observed a spontaneous learning of letters-order rules and no learning of colours-order rules. We interpreted these results as spontaneous attention oriented aspect of learning when it is more natural to learn letters strings rather than colours strings.
To verify this assumption in second experiment we used Arabic letters covered by different colours. Arabic letters are not familiar to westerners so it was possible to find out if participants would still be learning letters rather than colours. We didn't motivate people to learn any aspect of the material. We didn't observe the artificial grammar learning process in any condition.
In third experiment we used material from second experiment and we used three different instructions to direct participants' attention for learning one of the dimension — colours or letters. In control group participants were not focused on any dimension of stimuli. Likewise in previous experiment participants did not learn in any of three conditions. Such results provided new hypothesis combined with Eitam, Schul and Hassin (2009) observation. Attention engagement seems to be crucial during applying, not only acquiring tacit knowledge. Further experiments are needed to verify such assumptions.
UDC 159.95
N. Sidarus, P. Haggard
HOW DOES ACTION PRIMING INFLUENCE THE SENSE OF AGENCY?
Nura Sidarus — Ph.D. student at Institute of Cognitive Neuroscience, London, United Kingdom, n.sidarus.11@ucl.ac.uk
Patrick Haggard — Professor of Cognitive Neuroscience at Institute of Cognitive Neuroscicence and Psychology Institution, University College, London, United Kingdom, p.haggard@ucl.ac.uk
Sense of agency (SoA) refers to the feeling that we are in control of our own actions and, through them, of events in the outside world. One influential view claims that the SoA depends on a retrospective matching between the expected and actual outcome of an action. However, recent studies have revealed an additional, prospective component to the SoA, related to the fluency of action selection. The present study aimed to clarify the neural mechanisms of this prospective mechanism, by means of event-related potentials (ERPs). Participants responded to imperative stimuli that were preceded by either a compatible or an incompatible subliminal prime. After a variable delay, action outcomes were displayed, and subjective agency ratings were collected.
Results show that compatibly primed actions led to a stronger SoA over action outcomes, relative to incompatibly primed actions. At the neural level, target-locked lateralised readiness potentials were modulated by action priming, and were also positively associated with subjective agency ratings in compatibly-primed trials. Action priming did not influence outcome-locked ERPs. Thus, replicating previous studies, we found that an unconscious influence on action selection processes can affect the conscious experience of agency. These findings suggest that a metacognitive signal related to action selection fluency can prospectively inform the SoA. This signal is related to activity in the brain's
motor system. Furthermore, the influence of this prospective, fluency-based, component on SoA is independent from retrospective outcome monitoring.
UDC 159.95 A. Chetverikov
AFFECTIVE EVALUATION AS A MARKER OF IMPLICIT KNOWLEDGE
Andrey Chetverikov — Ph.D. student, Researcher at St. Petersburg State University, St. Petersburg, Russia, a.chetverikov@psy.spbu.ru
How can we understand that a person has acquired some knowledge? The most direct way is of course to ask "Well, have you learned something?", but as is well demonstrated by the studies of implicit learning this direct approach has its limits. Thus, a question of the most appropriate measurement arises. In this talk I will discuss affective evaluation ratings, such as liking, attractiveness or pleasantness, as a way of measuring knowledge. In their seminal work, Kunst-Wilson and Zajonc (1980) demonstrated that people like previously presented stimuli more than novel, even when recognition is at chance level. Further studies confirmed this finding (Bernstein, 1989). Not only the presented stimuli are preferred to novel, but stimuli that only resemble previous ones are more rated more positively as well. "Diffusion" of positive affect to novel stimuli associated will old ones is observed for prototypes (Winkielman, Halberstadt, Fazendeiro & Catty, 2006), stimuli from the same category (Monahan, Murphy & Zajonc, 2000), and even to stimuli that abide to implicitly learned rules (Gordon & Holyoak, 1983; Loui, 2012), although the latter effect has its boundaries (Newell & Bright, 2003; Zizak & Reber, 2004). Our own studies demonstrate that the affective effect of previous exposure may depend on the "decisions" in the accompanying task, such as whether stimuli are recognized or not, or whether they are classified as targets or distractors in visual search (Chetverikov, Johannesson & Kristjansson, 2014; Chetverikov, 2014). We suggest that affect can be regarded as a fringe feeling, such as a "feeling of rightness" (Mangan, 1993), indicating the accuracy of top-down predictions tested using available knowledge even in absence of explicit knowledge. However, there is a lot of unresolved questions regarding the relationship of affect with other measurements, such as confidence, and the factors that limit its usefulness.
The study is supported by RFHgrant 12-36-01294a2.
UDC 159.95 A. Lebed, S.Korovkin
CONSCIOUS & UNCONSCIOUS MENTAL PROCESSES COMPETITION: IMPLICIT LEARNING & CREATIVE THINKING VERSUS DECISION MAKING & ANALYTICAL THINKING
Anton Lebed — Ph.D. student at Yaroslavl State University, Yaroslavl, Russia, gyfest@yandex.ru
Sergey Korovkin — Associate Professor at Yaroslavl State University, Yaroslavl, Russia, korovkin_su@list.ru
Studies of thinking and particularly of problem solving are strongly hindered by researcher's inability to observe these processes directly. There are several methods using
indirect measures to be able to investigate hidden thinking processes more precisely. One of them is a dual-task paradigm. The idea of competition between mental processes for working memory resources is not new. For example, working memory impedes implicit learning (Reber & Kotovsky, 1997). This study aims to investigate the relation between several conscious processes (analytical problem solving, decision making) and unconscious ones (creative problem solving, implicit learning). We used a modified dual-task paradigm monitor subjects' working memory to elucidate the dynamics of problem solving (Korovkin, Vladimirov & Savinova, 2012, in rus). This method provides subjects with two tasks performed simultaneously (one primary and one probe-task of different modalities), while accuracy and reaction times in the secondary (probe) task is measured. Solver's limited resource is assumed to be withdrawn by primary task when there are more mistakes and longer RTs are in probe-task. However, there are several doubts considering this method. Most of them are about the validity of using working memory (which is supposed to serve as a resource for conscious control) to measure unconscious processes like insight problem solving or implicit learning. Some authors claim that working memory is the construct that is opposite to unconscious processes implying inability of using such methods (Dijksterhuis & Nordgren, 2006). We hypothesized that there is a competition between two unconscious processes (creative problem solving + implicit learning) and between conscious and unconscious ones (decision making + implicit learning). The obtained findings show that working memory monitoring within dual task paradigm can be used for investigating the dynamics of unconscious thinking processes. Additionally, we suggest that creative thinking and implicit learning are mostly functioning aside from decision making and analytical problem solving and probably have separate mechanisms. This suggestion is supported with rather limited interaction and competition of these processes during experimental tasks.
Present research is supported by Russian Foundation of Basic Research (12-06-00133a) and by the Grant of the President of the Russian Federation (MK-4625.2013.6).
UDC 159.95
I. Ivanchei, N. Moroshkina
HOW DOES CONSCIOUSNESS USE UNCONSCIOUS KNOWLEDGE? WHAT DECISION CRITERION CAN TELL US ABOUT LEARNING?
Ivan Ivanchei — Ph.D. student at St. Petersburg State University, St. Petersburg, Russia, i.ivanchei@ spbu.ru
Nadezhda Moroshkina — Assistant Professor at St. Petersburg State University, St. Petersburg, Russia, moroshkina.n@gmail.com
Accuracy of letter strings classification is the typical measure of the artificial grammar learning. We explored the decision criterion (how many strings are classified as grammatical). In signal detection theory conservative bias (low percentage of YES answers) is usually considered as rigorous strategy of decision making. We suggested that decision criterion can reflect the level of conscious control in artificial grammar learning.
In the first experiment we assessed the effect of conflict in upcoming information on the decision criterion. Conflict resolution is considered as one of the main functions of
consciousness in several theoretical approaches (e.g., Allakhverdov & Gershkovich, 2010; Botvinick et al., 2001; Frensch et al., 2003). We introduced conflicting stimuli in a standard AGL-experiment and observed the rise of the decision criterion.
In the second experiment we used the same design but added structural knowledge attribution procedure (Dienes & Scott, 2005) to explore the process more deeply: in each trial subjects had to report the basis of their classification decision. We failed to observe the influence of conflict on subjects' attributions but unexpectedly we observed the effect of attribution procedure on the decision criterion. Considering identical designs in two experiments we can conclude that the need to introspect the previous decisions led to an increase of the decision criterion.
We also made some additional measurements of conscious control relating them to between subjects variation of decision criterion.
The results show that the decision criterion can reflect subjects' level of conscious control over the task performance. People tend to classify AGL-items as grammatical rarely if conflicting stimuli are present and if they are forced to introspect their decisions. The second result also relates to the important issue of the "on-line" awareness measures: making decisions about ones performance in every trial can affect the decision strategy, i.e. the measurable behavior.
Supported by SPbSU (grant 8.38.287.2014).
UDC 159.95
F. Kemeny, A. Lukacs
SEQUENCE IN SEQUENCE: LEARNING OF AUDITORY BUT NOT VISUAL REPETITIONS WITHIN A MULTIMODAL SEQUENCE
Ferenc Kemeny — Senior Lecturer at BME Department of Cognitive Science, Budapest, Hungary, fkemeny@cogsci.bme.hu
Agnes Lukacs — Associate Professor at BME Department of Cognitive Science, Budapest, Hungary, alukacs@cogsci.bme.hu
While modality dependence in sequence learning has been a widely researched issue, the literature is very poor on the acquisition of multimodal sequences. The current design uses a modified version of the Serial Reaction-Time task. Participants were asked to solve two tasks simultaneously: one with auditory stimuli and one with visual stimuli. Their aim was to press the button that corresponds to the presented auditory or visual stimulus. There were four possible visual and four possible auditory stimuli, and both tasks used the same set of four response buttons. Unknown to the participants, there were a number of sequential patterns in the task: 1) a 12-element repeating response sequence, 2) a 12-element stimulus sequence of auditory items, 3) a 12-element stimulus sequence of visual elements, and 4) items of the visual and auditory sequences formed a fixed 24-element multimodal sequence where a maximum of three consecutive elements were from the same modality.
Participants were trained on seven blocks. Items from one modality were randomised in the eights, and items from the other were randomised in the ninth block. In block 10 the whole of the stimulus sequence was randomised. Results showed that RTs increase due to the randomisation of either part of the sequence.
At the end of the task participants faced a two-block unimodal testing: participants were assigned randomly into two groups, Group1 had auditory, Group2 had visual final testing. In block 11 participants faced one block of testing on the unimodal 12-element sequence that was part of the previous 24 element sequence, while in block 12 the sequence was removed, and elements appeared in a random order. Results showed that in the case of the visual sequence, RTs of the sequenced and random blocks did not differ from each other, while with the auditory sequence participants showed higher RTs for the random than for the sequenced block. This result shows that the visual items were only integrated into the multimodal sequence, while auditory elements also formed an auditory sequence.
UDC 159.95
V. Gershkovich, P. Iamshchinina
IS THERE CHOICE BLINDNESS EFFECT, IF CHOICE IS BASED ON IMPLICIT KNOWLEDGE?
V. Gershkovich — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, valeria. gershkovich@gmail.com
P. Iamshchinina — researcher at St. Petersburg State University, St. Petersburg, Russia iamshchinina@ gmail.com
The study was based on the hypothesis by Nisbett and Wilson (1977) that implicit attitudes that guide human behavior are stable and can't be distorted by conscious situational factors. Based on this assumption it might be supposed that choice based on implicit knowledge cannot be distorted by presenting misinformation about it and its verbalization as it happens within choice blindness situation (Jonhansson et al., 2006). In our experiment the situation of choice was modeled: subjects were presented with reproduction paintings and their mirrored variants with the instruction to choose either the original one (group N 1) or the one that seemed more attractive to them (group N 2). The stimuli were landscapes which contained rivers and roads directed from lower left to upper right corner. Group N 1 was previously informed about the rule that could help to differentiate reproductions; group N 2 received several examples of the rule, but was not informed verbally about the rule itself. There were then three stages of the experiment. At the first stage subjects made their choice. At the second stage subjects in both groups were presented with the pictures from the first stage and were asked to give explanation to some of their choices. However not chosen variants were presented as well though subjects didn't anticipate that. At the third stage subjects were asked to make choices among previously presented and new landscapes once again. During analysis we compared groups with implicit and explicit learning concerning the ability to remember choices after misinforming. We found that explicitly learnt group applied the rule in both previously seen and new pictures of landscapes, thus they were likely to return to their previous choices and misinformation effect was minimized. Second, we found that implicitly learned group was highly effective in applying the rule to new pictures but failed to return to their choices after misinformation. Although implicit learning results in some growth of efficiency as a whole it can be concluded that implicitly made choices are not stable and only highly deliberate choosing can withstand misinformation effect.
The study is supported by RFHS 12-36-01342-a2.
UDC 159.95 M. Kuvaldina
UNCONSCIOUS REACTION TO THE "UNSEEN" ITEM IN A SUSTAINED INATTENTIONAL BLINDNESS TASK
Maria Kuvaldina — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, kuvaldinamara@gmail.com
Inattentional blindness is the inability to notice an extra item while you attention is engaged in some other task. The main research question is whether we don't see this extra item (Mack & Rock, 1998) or whether we notice it and then tend to forget it (Wolfe, 1999). To verify the idea of inattentional amnesia some unconscious processing of the extra item should be recorded. Two main approaches were used: one is measuring deviations in performance during inattentional blindness task and the other — measuring the influence of inattentional blindness effect on subsequent task performance. Both methods have methodological complications that arise from the problems of detection of conscious versus unconscious information processing (Simons, Hannula, Warren & Day, 2007). A series of experiments has shown that the conclusion on whether the extra item is processed or not is dependent on the type of the measurement that was used. This result explains discrepancy in the literature on inattentional blindness (Bressan & Pizzighello, 2008; Most et al., 2005 and others) and leads to the discussion of the difference between inattentional blindness reaction and orienting response (Sokolov, 1960).
The study is supported by Grant of President of Russia No. MK-6068.2013.6.
UDC 159.95
Estibaliz San Anton, J. Bertels, T. Gebuis, A. Destrebecqz
DOES LEARNING MODULATE VISUAL ATTENTION IN INFANCY? AN EXPERIMENT USING THE HEAD TURN PREFERENCE PROCEDURE
Estibaliz San Anton — Ph.D. student at Center for Research in Cognition and Neurosciences (CRCN), Université Libre de Bruxelles, Belgium, esananto@ulb.ac.be
Julie Bertels — Postdoctoral Researcher at Center for Research in Cognition and Neurosciences (CRCN), Université Libre de Bruxelles, Belgium, jbertels@gmail.com
Titia Gebuis — Research Fellow at KU Leuven, Belgium, Titia.Gebuis@psy.kuleuven.be Arnaud Destrebecqz — Assistant Professor at ULB, Bruxelles, Belgium, adestre@gmail.com
One of the main learning mechanisms in infancy consists in extracting the statistical regularities present in the environment. The contextual cuing paradigm has been adapted to determine whether 5 to 13 years old children are able to learn the covariation between a set of distractors and the spatial location of a visual target (Couperus, Hunt, Nelson & Thomas, 2011; Dixon, Zelazo & De Rosa, 2010; Vaidya, Huger, Howard & Howard, 2007). So far, results in children are scarce and controversial. Moreover, this form of learning has not yet been studied in infancy. In our study, we modified the usual paradigm in order to make it more appealing and suited to young infants from 8 to 12 months of age. Following an habituation phase, we used the "head turn preference procedure" to measure the infants'
interest for spatial configurations that respected or not the association between a target location (a comics' character) and an arbitrary context of geometrical figures (Old vs. New configurations, respectively).
Results show that our participants (n = 38) were able to learn these regularities as they showed a familiarity effect. They looked longer at Old than New configurations. The same pattern of results was replicated when the context was made up of four or eight geometrical figures, suggesting that the size of the display does not influence learning. These findings suggest that the learning mechanisms subtending contextual cuing are already at work in infancy. We discuss the potential implication of these findings regarding the development of the visual orientation of attention.
UDC 159.95
Ryan B. Scott, Anil K. Seth
UNCONSCIOUS REWARD SPEEDS CONSCIOUS ACCESS WITHOUT CONSCIOUS PREFERENCE
Ryan Scott — Lecturer at University of Sussex, Falmer, United Kingdom, r.b.scott@sussex.ac.uk
Anil Seth — Professor of Cognitive and Computational Neuroscience, and Co-Director, Sackler Centre for Consciousness Science, Department of Informatics, University of Sussex, Falmer, United Kingdom A.K.Seth@sussex.ac.uk
It is established that the salience of a stimulus reliably affects the speed with which it is given conscious access; for example, highly familiar or emotional stimuli break through continuous flash suppression more rapidly than unfamiliar or emotionally neutral stimuli (e. g. Stein et al., 2012). Here we asked whether this preferential access extends to stimuli which have never been consciously perceived but which have acquired "unconscious salience". We first manipulated unconscious salience by pairing novel subliminal shapes with either reward or non-reward. We then compared the rate at which those shapes broke through continuous flash suppression with that of shapes not given prior subliminal exposure. Shapes subliminally paired with reward (M = 3876, SE = 412), broke through more rapidly both than shapes paired with non-reward (M = 4649, SE = 586), t(54) = 2.23, p = .030, and shapes given no prior exposure (M = 4552, SE = 546), t(54) = 2.64, p = .011. Shapes subliminally paired with non-reward did not differ in break through times from those not given prior exposure, t(54) = 0.43, p = .667. Interestingly, while unconsciously acquired salience accelerated conscious access it had no effect on the relative valence of the stimuli; the rated pleasantness of shapes subliminally paired with reward (M = 264, SE = 12), neither differed significantly from those subliminally paired with non-reward (M = 270, SE = 14), t(54) = .33, p = .664, nor those not given prior exposure (M = 263, SE = 12), t(54) = 0.07, p = .946. These results advance our understanding of the relationship between salience and accelerated conscious access, demonstrating for the first time that faster access extends to stimuli that have acquired salience without conscious perception and which do not differ in consciously perceived valence.
UDC 159.92
A. Y. Agafonov, Samara, Russia NECESSITY OF QUALIA
Andrey Y. Agafonov, Professor, Head of the Department of General Psychology, Samara State University, Samara, Russia, agafonov@ssu.samara.ru
The solution of the consciousness problem implies an explanation of experience nature (qualia), including how the brain activity is capable of causing subjective experience. However, the consciousness problem isn't just explanatory gap problem. Most likely there is no single "hard" consciousness problem. Other problems associated with the consciousness phenomenology are also important. I distinguish two problems concerning the necessity of qualitative experience. 1. What is the awareness functionality in the cognition process? Why do we need qualia, if the majority of mental processes and operations perform efficiently in the unconscious psyche «darkness»? Experimental facts rouse (us) to look for the answer, exposing to audit existing approaches. 2. Is there an objective need for subjective evidence of conscious experience? This problem implies an explanation of the objective necessity of qualia. Here is the basis for solving these problems. In the cognitive system, consciousness has a special role — testing hypotheses about the external world and awareness of the results. Whole mental product (ideas, interpretations, understanding patterns) is initially generated by special part of mind — «cognitive unconscious». «Cognitive unconscious» simultaneously builds several mental projections of the macroworld. The necessity of existing in a given reality requires overcoming redundancy of unconscious psychic projections. In one moment the consciousness checks one of the simultaneously existing unconscious hypotheses. Various conscious states neither logically nor empirically could be experienced simultaneously, so qualia at the current moment must be unique, and the change of qualitative states must occur sequentially, because becoming aware of reality in different ways simultaneously is impossible. Thus, conscious experience is the result of the choice and the check of the one of numerous unconsciously generated ideas. What can consciousness do and unconscious could not, though it seems being able to do everything? Unconscious cannot eliminate errors. The functional purpose of consciousness is the error realization, which allows further elimination.
Supported by RFBR grant № 13-06-00416 and by RFH grant No. 12-06-00457.
UDC 159.92
A. A. Kotov, T. N. Kotova
PRESCHOOLERS CONCEPTUAL FLEXIBILITY AS A FUNCTION OF IMPLICIT AND CONCEPTUAL LEARNING
Alexey Kotov — Research Fellow at Higher School of Economics, Moscow, Russian Federation, al.kotov@ gmail.com
Tatiana Kotova — Senior Researcher at Cognitive Research Centre, RANEPA, Moscow, Russia, tkotova@ gmail.com
The numerous contemporary cognitive studies reveal that the children and adult category learning involve relying on different kinds of context information (Macario, 1991;
Allen, Brooks, 1991). That effect was called the conceptual flexibility and became one of the central one in the fundamental leaning mechanisms studying in last ten years, because it makes new category learning very fast — the information assimilated in one context can be used in different conditions. The conceptual flexibility effect contradicts the most of the classical conceptual learning models, which maintain that by increasing the knowledge about the new category the attention to the relevant features is to be increasing, and the attention to irrelevant ones are to be decreasing in accordance with the cognitive economy requirements. While trying to remove this contradictory the researchers have found some important facts. Thus there were revealed that implicit and attentional learning may give rise to conceptual flexibility effect among 4-year-olds (Sloutsky, Fisher, 2008).
In our experiment we defined the conditions of conceptual flexibility development. We have made the new method ("multiple context categorization task") for 4-year-olds and adults conceptual flexibility estimation. This task can help us to find the age and conditions which make conceptual flexibility depended on implicit learning or conceptual learning. Our research has demonstrated that in those situations when subjects (both children and adults) couldn't find an appropriate cognitive scheme they increase conceptual flexibility only by the associative learning.
Multiple context categorization task can help us to find the age and conditions which make conceptual flexibility depended on implicit learning or conceptual learning. The conditions which influence we are going to investigate and present are: contextual information structure relevance to the learning category structure and adult's verbal signs for the contextual information.
This study was carried out within The National Research University Higher School of Economics' Academic Fund Program in 2014-2015, research grant No. 14-01-0168.
UDC 159.94
E. V. Bobrova, V. A. Lyakhovetskii, I. N. Bogacheva
EXPLICITNESS STAYING IMPLICIT: RIGHT OR LEFT HAND AND HEMISPHERE
Elena Bobrova — Leading scientific researcher at Pavlov Institute of Physiology, Russian Academy of Science, St. Petersburg, Russia, eabobrovy@gmail.com
Vsevolod Lyakhovetskii — researcher at Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia v_la2002@mail.ru
Irina Bogacheva — researcher at Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia boiss@mail.ru
Motor sequence learning by the right versus left hand was studied. Blindfolded righthanders had to repeatedly memorize and immediately reproduce the sequence of six right / left hand movements on sensory monitor. The explicit rule postulates that while the positions of the elements remain constant, the order of the elements changes in repeated probes. It is shown that the rule is used only in the case when the task is performed by the left hand, but not by the right one: left hand errors became lower in repeated probes. It is supposed to be explained by the inter-hemispheric differences in movement control and right hemisphere's specialization in positional (absolute) spatial coding versus left hemisphere's specialization in vector (relative) coding (Haaland e.a., 2004, Lyakhovetskii, Bobrova, 2007). The task performed by the left hand activates the low-level implicit representation of motor program
in the right hemisphere more than the task performed by the right one, and thus the rule about positional constancy is used. The possibility of using the explicit rule supposedly depends on coincidence or difference between high-level representation of explicit rule and low-level implicit representation of motor programs: the rule information is used in case of coincidence, but not in case of difference. The data are in agreement with singlesystem view (Cleeremans, Jimenez, 2002). Thus the explicit rule can be either used or not depending on the type of representation specific for hemisphere preferably activated during motor sequence learning.
Supported by RFBR grant No. 13-04-01071.
UDC 159.95 S. N. Burmistrov
LEARNING EFFECTS DURING THE ACTION OF DIFFERENT TYPES OF FEEDBACK
Sergey Burmistrov — Teaching Assistant at Samara State University, Samara, Russia, burm33@mail.ru
This research has been aimed at the identification of the effects of different types of feedback on cognitive performance. The research consisted of two experiments.
The first tested the assumption that feedback, informing the subject of a greater magnitude of the error than actually committed errors, can be an effective resource of correcting erroneous actions and double errors. In the experimental procedure, examinees performed the task of "firing of the gun". Required to hit the conditional goal which changed its localization. Results confirmed the initial hypothesis: learning effects are more pronounced when the feedback is false. There were also less double errors in this experimental condition compared to situation when the real feedback was given.
The second experiment tested whether the efficiency of learning depends on the contradictory feedback. We used the tasks of visual-motor coordination. A dot that moved along the circumference at a constant speed was presented on the screen. After passing a certain point of the circle the movement became unseen. The task was to stop the motion of the dot in the initial position, i. e. after passing through the entire circumference. After each trial two possible solutions were presented to the examinee: one reflected the actual result, the other — false. It was found that under conditions of contradictory feedback there is a strong tendency to accept the corrected result of two presented alternatives for their own. Usage of this strategy increased the efficiency of learning.
The above mentioned research demonstrated the potential of false and contradictory feedback usages during the process of learning (the potential for double mistakes correcting in the process of visual-motor coordination tasks fulfilling in particular). Furthermore, it was found that the double errors are important in the learning process, providing a more precise repetition of reactions.
The study was performed within the confines of a project supported by RFBR (grant No. 13-06-00416).
UDC 159.95
A. Y. Koyfman
SEMANTIC PRIMING: DISAPPEARED BEHIND THE MASK
Alexandra Koyfman — Psychologist at Moscow State University, Moscow, Russia, skoyfman@gmail.com
Unconscious semantic priming is a well known procedure of speeded presentation of a masked prime prior to target (both can be either written/spoken words or pictures seman-tically related to each other), that affects speed and/or accuracy of semantic categorization, lexical decision or naming task performance relative to baseline, where prime and target are unrelated. The amount of successful publications is impressive so it may seem that you only need a prime, a target and a mask in appropriate sequence to establish reliable priming-effect. On the other hand, you'd also need to set an appropriate timing to get in between objective and subjective thresholds: subject should perceive the prime and in the same time he or she should be unaware of that perception. Conditions vary a lot: you may obtain priming at 17-50 ms prime duration, so at first sight finding subjective threshold has higher priority.
There is a type of semantic priming where target represents an exemplar of category used as a prime — category priming (i. e. animal-lion). Due to semantic nature of stimuli used in semantic priming procedure there're many other variables to control along with timing. In several experiments different categorization tasks and categories, SOAs and masks were tested in attempt to achieve the category priming. They all resulted in failure. Why? Is the effect so fragile or whether the nature of problem is perceptual, but not conceptual? Final prime visibility measure showed that masking was effective and even too effective, so that prime was totally wiped off by the mask and fell below the objective threshold.
The results revealed a strong need of new "starting point" procedure to find an appropriate prime/mask durations with least possible costs.
Supported by RFBR grant No. 13-06-00053.
UDC 159.92 Zh. Gorkaya
IMPLICIT COMPONENTS OF INDIVIDUAL'S VALUE ORIENTATIONS SYSTEM
Zhanna Gorkaya — Associate Professor at Samara State University, Samara, Russia, zhanna_v_@mail.ru
A modern person is used to make important decisions in situations of obvious conflict of social values. Our consciousness creates multidirectional and sometimes even contradictory system of values preferences. Values are reflected in consciousness as worldview orientations that determine life goals and strategies of their achieving.
We were interested in researching implicit values and their correlation with explicit system of personality values orientations.
In our research we used methods of reflexive conformity value choices (Petrovsky V. A., 2006) and modified psychophysiological test of tension peak (Keeler, 1930).
In the first phase participants had to:
make two choices of values: a meta — choice normative values of the reference group and personal choice;
verbally classify values on to "important — not important";
rate their confidence in given response;
pass the psychophysiological test of tension peak.
Multichannel psychophysiological study shows the galvanic skin response (GSR), photoplethysmogram (PPG), respiratory amplitude and heart rate. It provided help to determine the degree of awareness of the claimed values.
The study distinguished group of people with low and high conformity values. Extent to which was determined on the basis of regulatory overlap and personal choices subject.
Analysis of the psycho-physiological results allowed to create the following groups of values (n=370):
1. Explicit (realized) values — stimuli that have identical verbal and psychophysiological meaning. They also can be differentiated on the accepted and rejected stimuli.
2. Implicit (not realized) values — stimuli that have differences between verbal and psychophysiological meaning. Implicit values influence on decision making and can be differentiated on the significant and non-significant.
At high conformity values of interpersonal coherence value preferences implicit value coincide with explicit (n=67, r=0,46, p<0,05).
At low conformity values of interpersonal coherence values implicit values do not coincide with the personal choice of the subject (n=102, p<0,01).
Thus, the implicit component values, influencing decision-making, can be determined by method of psychophysiological testing.
UDC 159.95 N. S. Kudelkina
DYNAMICS OF UNCONSCIOUS COGNITIVE SET EXTINCTION WITH AND WITHOUT REALIZABLE FEEDBACK
Natalya Kudelkina — Associate Professor at Ural Federal University, Ekaterinburg, Russia, kudelkinans@ gmail.com
Introduction. At unconscious level cognitive system "works" with series of unconscious influences (not with single stimuli) (Whittlesea, Jacoby 1990; Smith 2001; McKoon, Ratcliff 1995; Bodner, Masson 2001; Kiefer 2007; Agafonov, Kudelkina, 2010, etc.). One of the basic principles of the cognitive system work (including working with unconscious information) is prediction of each following impacts, based on previous experience. The cognitive system always has a prior orientation, a predisposition to percept and to respond in a specific way. So, (1) the degree of influence and (2) the nature of influence and (3) the fact of the presence or absence of influence of each following stimulus on the current conscious cognitive activity by the time of its occurrence is already set on the basis of the analysis of the previous series of influences. If we want to predict the effect of the stimulus, it does not matter to us what kind of stimulus we have, it's much more important what kind of stimuli we had before. These features of the cognitive system functioning entail the main problem of the research method — the 'null point' problem. How can we analyze the cognitive system's predispositions if they come from unique prior experience (we can't control this variable)? How can we make a «reset» of the individual cognitive system predispositions before the experiment?
In the experiments, presented here, we propose to use unconscious experimental stimuli which initially didn't render any noticeable influence on the effectiveness of tasks solution (neutral stimulus) and then to generate artificial settings for this stimuli by special experimental environment (it's important also that the unconscious stimuli aren't deduced on the perception level throughout all the experimental series). In this case, these artificial settings will be the same for our subjects. So, we can eliminate the "error" associated with the individual settings of cognitive systems. We offer to use serial (multiple) unconscious stimulation and the modified masked priming technique to give unconscious stimuli to our participants. Our experimental design is a combination of the masked priming paradigm and implicit learning paradigm.
Experiment. The goals: (1) to get steady priming-effect for the initially neutral unconscious stimuli, (2) to analyze the dynamics of this effect, namely, the dynamics of extinction of the priming-effect of unconscious stimulus (in experimental conditions with and without realizable feedback).
We hypothesized that the unconscious cognitive set will not extinct, if the subjects does not receive the realizable feedback about its inefficiency (mechanism of self-support for implicit settings).
Experiment procedure. There was developed the computer program (itis looks like a computer game for our participants).
Target stimulus: 1) "puzzles" — pictures preliminary cut on 9 parts and mixed in random order. In general we use 24 pictures, belonged to the same artist. We select puzzles of approximately identical level. Using puzzles, we created experimental tasks of two types: (a) «Solvable puzzles» — it was possible to put together an original picture; (b) «Unsolvable puzzles» — it was impossible to put together the complete image. «Unsolvable puzzles» were received by replacement of three fragments of a picture by their mirror reflections.
Primes stimulus: The graphic objects of two kinds — it's look like «X» and «Z». These symbols initially didn't render significant influences on the solution of the subsequent tasks. It was checked in a pilot study beforehand.
Plan of the experiment: The preliminary series consisted of 10 tasks. All the solvable tasks of a preliminary series were accompanied by unconscious prime «X». All the unsol-vable — by prime «Z». The sequence of solvable and unsolvable tasks in the experimental series was set in a random order. Then, there came the control series consisting of 14 tasks. All the tasks in this series were solvable, but half of them were accompanied by unconscious prime «X», the other seven tasks — by prime «Z».
Experimental procedure: The examinee should collect a complete picture as soon as possible, shifting by means of a computer mouse two fragments of a puzzle. In the process of collecting of a picture a prime («X» or «Z») was shown to the examinee. Conditions of prime presentation excluded possibility of its comprehension (presentation time 20 msec). The instruction to the examinee: «You should collect the pictures as soon as possible. Among the offered to you pictures there are pictures which can't be collected. If you are absolutely sure that it's impossible to collect the picture, press ESC. It will allow you to pass to the following puzzle. Please don't distract!»
We had two experimental groups. The participants of the first experimental group received no feedback during task solving. The participants in the second group received the feedback: if someone tried to mark the task as unsolvable, but in reality it was solvable puzzle, the message appears on the screen "Error! Try to solve again".
Results. (1) The tasks of a control series accompanied with prime «Z» (the prime «Z» earlier in a preliminary series was applied with "unsolvable" puzzles), were being solved less efficient, than similar tasks with prime «X» in both experimental groups. They required more time for solving. The examinees did more erroneous shifts of the fragments while solving the tasks with prime «Z» and the examinees more often refused to collect a puzzle with unconscious prime «Z» because of the subjective confidence of its «insolvability» in comparison with puzzles with prime «X».
(2) In the experimental group without feedback, we have not detected cognitive set extinction throughout the entire experimental series (14 tasks). Especially this result is expressed for negative-priming effect of Z-primes. The effectiveness of tasks solving with Z-primes, remained low for all of 7 tasks of the control series. However, in the group with feedback, the negative-priming effect of Z-primes extinct after 3-5 tasks of control series.
Discussion. During preliminary series, the cognitive system detects the implicit patterns (interrelation between the Z-primes and inability to solve the puzzles). Then, at the beginning of a control series, the cognitive system falls into a dissonance (the implicit pattern does not correspond to reality anymore). All of the puzzles in the control series are solvable (even with Z-prime). So, there is a choice: (a) to abandon the implicit pattern (to extinct the implicit cognitive settings) or (b) to make the reality corresponding to the settings. In the experimental group without feedback the participants have the opportunity to make the reality corresponding to the cognitive settings — they continue to refuse to collect the puzzles with unconscious prime «Z» because of the subjective confidence of its «insolvability». So, the tasks with Z-primes remain subjectively unsolvable until the end of the experiment. The participants of other experimental group were unable to «change» the reality (in accordance with the unconscious expectation) because of the feedback.
Conclusions. The implicit cognitive settings are quite resistant and inert: even if they don't correspond to reality anymore, the cognitive system tries to keep them by changing the perception of the reality according to the implicit settings (mechanism of self-support for implicit settings).
The results of our experiment is well correlated with the data published by T. Hill, P. Le-wicki, I. Sasaky. They report that encoding (interpretative) biases may develop in a self-perpetuating and self-supportive manner (even in the absence of any objectively supportive evidence).
UDC 159.95
S. S. Belova, E. A. Valueva
IMPLICIT LEARNING IN A FRAMEWORK OF INDIVIDUAL DIFFERENCES APPROACH
Sofya Belova — Research associate at Institute of Psychology, Russian Academy of Science, Moscow, Russia, sbelova@gmail.com
Ekaterina Valueva — Research associate at Institute of Psychology, Russian Academy of Science, Moscow, Russia, ekval@list.ru
The present study was aimed at application of individual differences approach to implicit learning (IL) in order to test two hypotheses. The first one implied existence of general factor of ability for IL across domains. The second one suggested that ability for IL correlates positively with general cognitive abilities, i. e. intelligence and creativity. Four measures of
learning were used, two of them being aesthetic (categorization of pieces of art according to their style and categorization of pieces of Japanese poetry according to their style), one being social (categorization of faces according to their configural types), and an artificial grammar task (AGT). Each measure contained a learning session which could be implicit or explicit and a testing session. The participants were randomly assigned to an implicit or explicit learning group (N=39, N=38 respectively, mean age 20.6). Accuracy of performance was used as a dependent variable. Four main results were obtained. Firstly, accuracy of performance was above chance in all measures and conditions except for AGT and the poetry task in IL condition. Secondly, factor analysis of the implicit group data revealed one factor explaining 35% of variation having all the measures loaded positively except for art style task which loaded negatively. Factor analysis of the explicit group data revealed one factor explaining 40% of variation having all the measures loaded positively (0.35-0.75). Thirdly, positive correlation of performance in explicit learning condition and intelligence was revealed whereas no correlation of performance in IL condition either with intelligence or with creativity was found. Fourthly, the specific pattern of correlations of intelligence and decision making criteria was revealed. Higher intelligence was associated with more frequent inclusion of a stimuli into a target category in the condition of IL of nonverbal regularities (in the art style and social perception tasks) whereas in the explicit condition the opposite was true. The obtained results are discussed within a dual-process framework of individual differences.
The study is supported by RFHgrant 14-36-01293a2.
UDC 159.95 N. Andriyanova
PREDICTING OF REGULAR ERRORS IN THE LEARNING PROCESS
Natalya Andriyanova — Ph.D. student at St. Petersburg State University, St. Petersburg, Russia, andriyanova89@mail.ru
The current study investigated predicting of regular errors in the learning process by means of reaction time and the level of confidence. It aims to describe the phenomenon of regular errors and to determine the features of wrong answers at the beginning of learning process which enable predicting of regular errors during the course of learning process. There are a lot of evidences on regularities of appearance and repetition of errors in the learning process. However such regularities are usually registered after completion of learning process and errors tend to consolidate. This work is directed on detection of regular errors at the initial stage of learning process that will allow to correct such errors in due time. It is generally assumed that reaction time is less for regular errors than for singular errors.
There are two experiments in the current research. In the first experiment on the computer screen 120 clock faces are shown to subjects for 200 ms, one at a time. After presentation of each stimulus subject needs to enter time which was shown on the clock face. Stimuli are grouped in 10 series; in each series 12 different indications are used. Current experiments demonstrated that reaction is faster for regular errors in the beginning of learning process, than for singular errors. This is the signal that such errors will be repeated in the course of learning process. Besides, a slowing of correct answers before regular errors was found. Confidence of the answer is also different for different types of answers. The level
of confidence of regular wrong answers increases in the course of learning process. The structure of the second experiment is the same, but we use chessboards as the stimuli and subjects should answer about the position of chessmen, and the results are the same too. The study is supported by RFHgrant No. 13-06-00535 a
UDC 159.95 V. Karpinskaia
THE ROLE OF PERCEPTUAL SET IN THE PROCESS OF DISCRIMINATION THE STIMULUS
Valeria Karpinskaia — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, karpinskaya78@mail.ru
We investigated the role of perceptual set in the process of discrimination the stimulus. Usually, in psychophysics the process of detection and discrimination depends on physiological factors and the physical parameters of the stimulus. But there are also psychological factors that can change the efficiency of these processes. One of it was shown in our previous experiments — illusory change of the stimulus influenced sensory thresholds (Karpinskaya, 2010, 2011, 2012). Perceptual set is the other way to change the size of the stimulus not objectively but subjectively. The patterns of Ponzo illusion was shown to the participants. The illusion was modified: there were two figures of the same size in the context with perspective instead of horizontal lines. In each pattern the figure, that looked bigger, was on the same place (left of right side). The session consisted of fifteen trials with the patterns, and after that, two circles of the same size were shown without any context, one of them was standard. The participants had the task of making two circles the same visible size. He had only 5 seconds to make a decision. And after that was a new session. Each participant had a shot training, 10 sessions with illusion, 10 sessions without illusion (different pictures of geometrical figures), 10 sessions with pictures of figures with difference in size. After each session participants had the task with two circles. Results showed that the process of discrimination was influenced by the previous trials that used Ponzo illusion or by the perceptual set in session with figures of different size. The mistakes in the size of the circles depended on the perceptual set.
The authors acknowledge St. Petersburg State University for a research grant 8.38.287.2014.
UDC 159.95
A. Karpov, N. Moroshkina
INTERACTION OF IMPLICIT AND EXPLICIT KNOWLEDGE IN SOCIAL PERCEPTION TASKS.
Anatoly Karpov — Ph.D. student at St. Petersburg State University, St. Petersburg, Russia, anatoly1804@ gmail.com
Nadezhda Moroshkina — Assistant Professor at St. Petersburg State University, St. Petersburg, Russia, moroshkina.n@gmail.com
Several experimental studies demonstrated that unconscious processes can play a crucial role in performing of social perception tasks, particularly when subjects evaluate
different psychological characteristics by appearance (Lewicki et al., 1986, 1987; Barker, Andrade, 2007). However, this effect was not reproduced in the later series of studies (Hendricks et al., 1997). In our experiments we investigated the interaction of implicit and explicit knowledge of the subjects in the evaluation of psychological characteristics of another person. In the first phase of the experiment subjects were presented against 20 pictures of females, under each photo there was value of IQ, which supposedly has this person. The task was to remember the faces of those persons whose IQ exceeded 100 points. The participants were not informed about a hidden covariation between the hairstyle and IQ level. In the test phase subjects were shown a series of new females photos with different hairstyles, and now the task was to evaluate their intelligence. In the third stage, participants were presented with photos of females from the test series, whose IQ had already been evaluated, and the subjects had to justify their choice. The results showed that people implicitly learned hidden covariation and this rule was applied to the evaluation of new persons at the unconscious level. The most interesting result is a negative correlation between the probability of the use of implicit knowledge and the ability to verbalize the reasons of their answers. Subjects, who had used the implicit criterion gave less detailed justification for their choice and often resorted to intuitive answers. Furthermore, it was found that such individual variables as impulsivity correlated positively with the use of implicit knowledge. However despite a lot of experimental evidence of implicit learning in social perception tasks (and the mentioned-above result included), we failed to observe the influence of hidden implicit covariation in series of our similar experiments. Thus the conditions necessary for its occurrence are still to be determined.
Supported by RFBR (grant No. 12-06-00311-a).
UDC 159.95
S. Korovkin, I. Vladimirov, A. Lebed, A. Savinova, A. Chistopolskaya
PROBLEM SOLVING AND WORKING MEMORY: RESULTS AND PROSPECTS
Sergey Korovkin — Associate Professor at Yaroslavl State University, Yaroslavl, Russia, korovkin_su@list.ru Ilya Vladimirov — Associate Professor at Yaroslavl State University, Yaroslavl, Russia, kein17@mail.ru Anton Lebed — Ph.D. student at Yaroslavl State University, Yaroslavl, Russia, gyfest@yandex.ru Anna Savinova — student at Yaroslavl State University, Yaroslavl, Russia, anuta1334@yandex.ru Alexandra Chistopolskaya — Ph.D. student at Yaroslavl State University, Yaroslavl, Russia, chistosasha@ mail.ru
The most promising direction in psychology of problem solving is the study of dynamics of insight problem solving. We assume that solving of insight and algorithmic problems involve different kinds of processes. However, the nature of these processes is largely unclear. The dynamics of problem solving cannot be directly measured. Nevertheless, one can monitor the dynamics through indirect indicators. One of the possible evidences of problem solving is the dynamics of working memory that can be monitored using probe task.
We were investigating the dynamics of the working memory in insight and algorithmic problem solving. The results have shown that the dynamics of the insight problem solving is significantly different in these two types of tasks. Our results suggest the presence of specific insight problem solving mechanisms (Korovkin, Vladimirov, Savinova, 2012, in rus.; Savinova, Korovkin, 2012, in rus.), insight problem solving is block-specific in working
memory (Vladimirov, Korovkin, Chistopolskaya, Savinova, 2013, in rus.), insight problem solving is partly performed in a special block of working memory which is responsible for unconscious processing (Lebed, 2014, in rus.).
Using the dynamics of working memory load as an indicator of insight problem solving has some limitations. First, the lack of a theoretical model of insight does not allow us to make clear hypotheses about the current mechanisms of problem solving. Second, the lack of a common understanding of the mechanisms and blocks of working memory significantly reduces the accuracy and adequacy of the probe tasks selection.
There are some points of growth of the dynamics of the working memory in problem solving studies. It is necessary to search a neural and brain basis for hypothetical problem solving mechanisms modelling. We should clarify an executive control's role in the model of working memory. These shifts of the theoretical models of problem solving and working memory will significantly increase the effect of research using the probe task techniques.
This work is supported by Russian Foundation for Basic Research (No. 12-06-00133a) and Grant of the President of the Russian Federation (No. MK-4625.2013.6).
UDC 159.92 T. Kotova, A. Kotov
THE GETTING OF CHILD'S ATTENTION TO THE ACT OF NAMING AS REFERENTIAL CUE AND AS JOINT EXPERIENCE INFORMATION
Tatyana Kotova — Senior Researcher at Cognitive Research Centre, RANEPA, Moscow, Russia, tkotova@gmail.com
Alexey Kotov — Research Fellow at Higher School of Economics, Moscow, Russian Federation, al.kotov@ gmail.com
In the natural circumstances a new word learning imply new object demonstration, new word pronunciation and getting of child's attention to the act of naming (GAN) simultaneously (Baldwin, 1991). The GAN means that adult implements naming intentionally with an appeal to child's consciousness, signs referential intention explicitly. That's why it has crucial importance in discussion about the role of joint attention in word learning (Achtar & Gernsbacher, 2007).
But the special role of GAN was not explored, because GAN is usually associated the new object demonstration and the new word pronunciation together.
Such procedures don't allow to understand, if GAN is a "label cue", and its function is to attract child attention to label, or an "invitation" to child-adult convention about object.
The goal of our experiment was to define function of GAN by dissociation new object demonstration and new word pronunciation.
We demonstrated new artificial object to 3-year-olds in four conditions. In the first condition we said: "I'll name this later!" and through 2-minutes-play we pronounced the artificial label (GAN accompanied new object demonstration). To second group we didn't gave any notification about further naming during object demonstration and through the same play we pronounced artificial label within phrase: "Now I name this! It is vepa!" (GAN accompanied new word pronunciation). Third condition was the combination of the first and second conditions (GAN accompanied both). In control condition we didn't attracted child's attention to the act of naming.
The main result was that only first and second groups performed mutual exclusivity task very well (used GAN as joint experience information), but they returned object for name significantly worse (didn't use GAN as referential cue). Thus we suppose that the function of explicit referential signs is not related only to label learning, it is convention making.
UDC 159.95
Ya. Ledovaya, K. Puzanova
HOW FIELD INDEPENDENCE RELATES TO IMPLICIT REGULARITY REVELATION
Yanina Ledovaya — Senior Lecturer at St. Petersburg State University, St. Petersburg, Russia, y.ledovaya@ psy.spbu.ru
Ksenia Puzanova — student at St. Petersburg State University, St. Petersburg, Russia
Although for the recent years there has been shown less interest to studying cognitive styles (controls), these notions can still help to reveal regularities in both explicit and implicit cognitive processes. Field dependence has been considered to reflect the level of perception control (Jonassen & Grabowski, 1993). There is much evidence about field dependence-independence (FD/FID) and task success in highly and lowly structured tasks: FID tend to excel FD in learning and recalling poorly structured information (Tannenbaum, 1982; Skaggs, Rocklin, Dansereau & Hall, 1990), being able to create virtual structure in the information provided.
It is also known that FID are often focused on socially relevant information (Witkin et al., 1979; Lu & Suen, 1995).
The implicit regularity revelation was a side effect in the study of emotional design objects' perception.
In the study aimed to discover how the level of FD (measured by Ettrich AKT-70 test) relates to subjective attractiveness of anthropomorphic and non-athropomorphic objects of industrial design 30 pairs of photographs of such objects were presented as PowerPoint presentation to 51 participants (20 males, mean age 27). The median FD value divided the group by two. Participants were asked to choose a more attractive item between the two each time (it could be furniture, dishes, other everyday objects with similar functions). They also commented their choice. After presentation they were asked if they had spotted any regularity in the series.
Besides the fact that FID chose anthropomorphic objects less often (40% vs. 53%, p = 0.005), they revealed the simple implicit regularity (anthropomorphism) more often (80% vs. 38%, p < 0.01). We consider this result as manifestation of a higher differentiation level within the cognitive system of FID (Witkin et al., 1979), and also as manifestation of not only their ability to structure, but also the ability to reveal implicit structures.
The authors acknowledge St. Petersburg State University for a research grant 8.38.303.2014.
UDC 159.95
M. Morozov, V. Gershkovich
THE RECOGNITION OF PARTIALLY AND FULLY PRESENTED IMAGES: THE ROLE OF GENERATION EFFECT
Maxim Morozov — student at St. Petersburg State University, St. Petersburg, Russia, 100club@mail.ru
Valeria Gershkovich — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, valeria.gershkovich@gmail.com
The visual system is adept at compensating for the missing information in scenes that results from occlusion, but how this is done is not fully understood (Johnson, 2005). Although our impressions of object wholeness could arise at a purely conceptual level, there are also reasons to think that our visual system has some mechanisms at the perceptual level to account for occlusion and other forms of missing information in the early stages of visual processing (Nakayama, He & Shimojo, 1995). We need less then 50% of object contour to identify it (Shelepin et al., 2008, in rus.). But it is also shown that matching the object with the prototype can negatively influence people memory about the object (Cabeza et al., 1999). Consequently, we assumed that completion of the partially occluded objects could lead to false memory mistakes of remembering completed objects instead of their occluded versions at a recognition task.
At the first stage of an experiment subjects were presented with complete and occluded images of common objects and were asked to identify them. At the second stage, which took place in a week, subjects were asked to recognize previously presented objects among new and old ones. Previously presented old objects were changed: several occluded images were changed to their complete versions, previously complete images were changed to occluded ones, the rest were not changed.
According to our results, participants effectively distinguish new stimuli from the old ones. Our hypothesis that subjects would falsely recognize completed objects instead of previously occluded ones was not verified. On the contrary subjects made significantly more mistakes, recognizing occluded images as previously shown them instead of complete ones. We suggest that those false recognitions are caused by generation effect taking place at recognition stage: participants generate occluded image to the complete one and it starts matching with the memory for this object as it was shown at the first stage.
The authors acknowledge St. Petersburg State University for a research grant No. 8.38.287.2014.
UDC 159.95 A. Muss
CATEGORIZATION AND INATTENTIONAL BLINDNESS: HOW CAN WE PERCEIVE THE UNCATEGORIZED STIMULUS?
Alexandr Muss — student at St. Petersburg State University, St. Petersburg, Russia, albertwanderer@ gmail.com
Nowadays there are a lot of scientific theories of innatentional blindness, but none of those can fully explain this phenomenon (Kuvaldina, 2010, in rus.). In order to make this
question clearer we tried to examine the effects of categorization process on inattentional blindness. We tested the following hypothesis: uncategorized stimulus cannot be perceived (Tchostov, 2002, in rus.). During our previous experiments (Muss, 2013, in rus.) we found no differences in detection rate between two unexpected stimuli: one that comply with the main task and one incongruous with that task.
In this experiment 40 participants (mean age — 19,6) with and without previous knowledge about inattentional blindness carried out 5 trials of dichotic listening task: they counted the number of target sounds at the right ear while fixating their gaze at the center of the screen. After several trials when only an audial stimulus was presented on 4th and 5th trials both visual and audial critical stimuli simultaneously were showed to the subject. After those trials they filled out the questionnaire (similar to those in Most et al., 2001) in order to indicate if they detected any differences and noticed one extra item.
We found that two extra stimuli were detected significantly more often than only one stimulus and more often than each stimulus separately (4th trial: x2(3; N = 40) = 15.2, p = .002); 5th trial: x2(3; N = 40) = 15.2, p = .002). Interestingly, when our participants tried to recognize extra stimuli they recognized visual critical stimulus significantly more often than audial one. According to V. Allachverdov's theory of consciousness (1993 in rus.) we can interpret our findings as a result of the incongruity of visual critical stimulus with the main task: if the stimulus is inconsistent with current task, information about this stimulus tends to remain conscious.
UDC 159.95
O. Naumenko, Z. Gagieva
DOES SEMANTIC SATIATION HAVE NONSPECIFIC INFLUENCE ON PERFORMING COGNITIVE TASKS?
Olga Naumenko — Postdoctoral Researcher at St. Petersburg State University, St. Petersburg, Russia, olga_naumenko@bk.ru
Zarina Gagieva — student at St. Petersburg State University, St. Petersburg, Russia
When the word is many times repeated it temporarily loses its meaning. This phenomenon is called semantic satiation. Usually semantic satiation is regarded as having specific influence on processing of the repeated word itself (Jakobovits, 1962).
There is an alternative view considering semantic satiation effect to be a particular state of consciousness (Baars, 2013). Thus, semantic satiation could have nonspecific influence on performing cognitive tasks. We assumed that semantic satiation could empty consciousness from repeated errors and decrease the number of right answers omissions.
In our previous experiments Ss were to judge time on rapidly presented dials or to solve dual-meaning anagrams.
In our present experiment Ss were asked to name pack of 36 playing-cards from memory in a random order at first try. There were few pauses during the main task. During the pauses Ss had to (1) repeat the same word (irrelevant to the main task) until it loses its meaning; (2) read aloud 15 different irrelevant words; (3) wait for 20 seconds silently. Some Ss (4) had no pauses at all. The number of right answers omissions and the RTs were measured.
Previously we found the increasing number or right answers in the task of discerning dials and also the speeding-up of finding the second word in dual-meaning anagrams when the same irrelevant word was being repeated during the pauses. In discerning dials task we also found influence of the pause itself, though it was weaker than the effect of "satiation" pause. In anagrams solving there was no condition of pause absence.
In our present experiment no differences between the groups were found.
Seemingly, semantic satiation could have nonspecific improving influence on performing tasks which include the possibility of setting and restructuring "gestalts" from the elements.
The research is supported by RFHS grant No. 13-06-00535a. UDC 159.95
M. Nelubov, V. Gershkovich
EFFECT OF DIVIDED ATTENTION ON FORMING OF FALSE MEMORIES
Matvey Nelubov — student at St. Petersburg State University, St. Petersburg, Russia, mateus64@yandex.
ru
Valeria Gershkovich — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, valeria.gershkovich@gmail.com
There has been considerable debate about the relationship between explicit and implicit processes in the forming of false memories (e. g., Kawasaki, 2006).
The experiment was conducted in which DRM-procedure was used for inducing false memories. DRM paradigm involves presentation of a list of related words and then requires the subject to remember as many words from the list as possible. Typical results show that subjects recall or recognize a related but nonpresented word (Roediger & McDer-mott, 1995). The divided attention method was used for explicit memory suppression. This technique involves performing of main task simultaneously with a distractor-task which implicates tracking the target number sequence among auditory presented numbers. It has shown its effectiveness in suppression of explicit memory (Jacoby et al., 1993). There were 4 experimental groups: with full attention, with divided attention at study (to estimate the encoding of information), divided attention at test (to estimate the retrieval), and divided attention at both stages. We were interested in specific influence of explicit processes, taking place on different stages, on true and false memories.
The results show an increase in false memories in all conditions with divided attention and the effect is stronger when attention is divided both at study and test. Also there is a reduction in true recognition when attention is divided at study, as compared with other conditions. In general, it is supposed that implicit memory supports false memories, and reliance on the explicit component is essential to avoid false memories, while encoding may appear to be a process that is more demanding to conscious control than recognition, yet this is a task for a future research.
The authors acknowledge St. Petersburg State University for a research grant No. 8.38.287.2014.
UDC 159.95
V. Spiridonov, A. Karnaukhova
IMPLICIT LEARNING IN TEXT COMPREHENSION
Vladimir Spiridonov — Director at Center of Cognitive Studies at Russian Presidential Academy of National Economy and Public Administration, Moscow, Russia, vfspiridonov@hse.ru
Alexandra Karnaukhova — Ph.D. student at Russian State University for the Humanities, Moscow, Russia, ekarnauhova@mail.ru
This research is dedicated to investigation of implicit learning (IL) on verbal material during solving tasks of comprehension and differentiation of text fragments. We assume that in this case IL is based on regularities which differ from those in perceptional tasks or in artificial grammar tasks where IL is likely to be related to accumulation of frequencies of superficial features or letter strings (Pacton et al., 2001; Shanks, 2005). Solving a semantic task on the basis of natural language cannot be restricted to processing of superficial features of the material. That's why we assume that in this case Il processes involve cognitive mechanisms responsible for processing of superficial grammar structures. Consequently, if we find IL in this research we will acquire arguments supporting implicit learning of rules with regard to dilemma «rules vs. statistics» (Cleeremans & Dienes, 2008).
According to our experimental design participants are required to differentiate fragments of text as "true" or "false" on the basis of a superficial grammar rule — presence of passive voice forms in "false" texts. In cognitive linguistics superficial active and passive voices is transferred to semantic structures by the means of semantic protoroles put forward by D. Dowty (Dowty, 1991) or a similar set of deep cases, proposed by Ch. Fillmore (Fillmore, 1968; Novikova, 2008).
According to our hypothesis, learning in this case may rely on activation of deep structures relevant for the task being solved. It can be found in form of IL or in example generation task (for example, agent constructions may prevail non-agent constructions). In this way participants will learn to react and to use grammar regularities embedded in text fragments without becoming aware of these regularities. Confirmation of this hypothesis will provide evidence to the notion that IL may be based on non-statistical regularities embedded in material.
UDC 159.95
I. Ovchinnikova, N. Moroshkina
IMPLICIT LEARNING IN PERCEPTIVE CATEGORIZATION TASK: DISSOCIATION BETWEEN VERBAL AND IMAGINATIVE REPRESENTATIONS.
Irina Ovchinnikova — Masters student at St. Petersburg State University, St. Petersburg, Russia, ovchinir@gmail.com
Nadezhda Moroshkina — Associate Professor at St. Petersburg State University, St. Petersburg, Russia, moroshkina.n@gmail.com
Based on the theories of dual coding (Paivio, 1986, etc.), we have assumed that in the process of learning the subjects formed two types of representations — verbal and imaginative. As experience increases, the level of coherence between those two types of representa-
tions leads eventually to the explicit understanding of rule. Consequently, in the case of implicit learning, we must find a lower coherence between verbal and drawing tests of knowledge, as compared to the group with explicit learning (Exp. 1), and the higher coherence as compared with the group without learning (Exp. 2). There were 2 experiments, participants were asked to categorize objects (the rule was not known), and feedback was given. They were presented geometric shapes, which varied several parameters, served as the basis for categorizing the conjunction of two features: the concavity and vertical symmetry. In the Exp. 1 the subjects were divided into consciously and do not realize the categorization rule (verbal test). Both groups of subjects have shown the effectiveness above chance, but the use of drawing tests revealed a number of differences between these groups at the level of figurative representations. The Exp. 2 was similar to the first, but ISI was reduced to the subjects rely heavily on intuition. Also we varied the value of feedback: one group of subjects was given congruent feedback; the other group was given incongruent feedback, according to the learning curve of the first group (i. e., feedback was not dependent on the correctness of the answers and only formed an "illusion of learning"). This was done to compare the quality of figurative representations in the group with implicit learning and group with illusion that they have learnt something. The results have shown that both groups estimated their efficiency higher than chance although only the first group learnt implicitly. Verbal test revealed no differences between the two groups (all of subjects figured out false signs), but drawing tests of knowledge showed that the degree of coherence of drawings and verbal hypotheses in the group with implicit learning still higher than in the group without learning, which confirms our initial hypothesis.
The authors acknowledge St. Petersburg State University for a research grant 8.38.287.2014.
UDC 159.95
A. Shtykhina, I. Vladimirov
INSIGHT PROBLEM SOLVING AS A RESULT OF UNCONSCIOUSNESS THOUGHT. THE ROLE OF NEGATIVE AFFECT IN CONSCIOUSNESS CONTROL SUPPRESSION
Anastasiia Shtykhina — student at Yaroslavl State University, Yaroslavl, Russia, anastasia.kelsina@yandex.ru Ilya Vladimirov — Associate Professor at Yaroslavl State University, Yaroslavl, Russia, kein17@mail.ru
Unconscious processes are considered to play an important role in the insight problem solving. One of the modern models is the Unconscious thought theory by Dijksterhuis, Nordgren (2006). The main position of the theory is that there are two modes of thought: conscious and unconscious, they work differently. The unconscious works bottom-up (automatically, without strong rules), the conscious works top-down (schematically, following strict rules).
The idea is that conscious capacity is limited, in some cases unconscious search is more preferable then the conscious one, that allows to overcome an impasse during problem solving. In the experiment participants who were distracted for a while were more likely to solve the problem; they call it the effect of set shifting.
The method is based on the disruption of cognitive functions and the activation of unconscious processes by distracting working memory.
Similar method was used by Kristjansson et al. (2012) in studying the priming of visual search. Their experiment involved presentation of visual distractors (neutral, emotion-evoking pictures, inverted version of first) before the task.
The significant method difference of our experiment is the use of visual distractors of two types — negative and neutral pictures. We show them several times during the problem solving process (starting at the tenth second and ending with the finding of the solution).
We decided to verify some principles of the unconscious thought theory — existence of two modes of thought and their interaction in case of alternating periods of conscious and unconscious thought during the problem solving situation.
The method consists of the use of visual distractors of two types and tasks of two types (insight and regular). We had four types of tasks. The model will be confirmed by checking whether the performance in negative visual stimuli trial is more efficient than with neutral ones in case of insight tasks.
This work is supported by Russian Foundation for Basic Research (No. 12-06-33008).
UDC 159.95
O. Vakhrameeva, A. Harauzov, Yu. Shelepin
UNCONSCIOUS VISUAL PRIMING EFFECT ON FOVEOLAR MECHANISMS OF OBJECT SEGREGATION FROM BACKGROUND
Olga Vakhrameeva — Researcher at Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia, olga.vakhrameeva@gmail.com
Alexey Harauzov — Senior Researcher at Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia, harauzov@gmail.ru.
Yuri U. Shelepin — Professor, the head of Laboratory of Visual Physiology I. P. Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia, yshelepin@yandex.ru
In previous studies correlation between foveola diameter and performance in discrimination task has been found. Subjects detected gap orientation in the Landolt C element of size smaller than 0.5 deg the better the larger foveola size they have.
To investigate the possibility of unconscious influence of information received by magnocellular stream of the visual system on perception of stimuli addressed to parvocel-lular stream.
We used 7 hiragana letters (from Japanese alphabet) as stimuli in match-to-sample task. The task was realised in EventIDE. In the middle of the monitor object-sample appeared for 150 msec, and then, after 1200 msec, test-event was presented for 400 msec. In test event 4 objects situated on the screen and using keyboard subject had to choose the one which was presented earlier as an object-sample. Object-sample size was 0.1 or 0.2 degree, and noise level was 0 or 40%; 900 msec after object-sample, object prime was presented for 150 msec. Object-prime appeared in four corners of the monitor simultaneously. Subject could not notice consciously presence of primer, because it was masked. In primer visibility test, when subjects had to attend to primer and identify it, performance was 83%. Each prime size was 4 deg, the distance between central objects and objects-prime was 6 deg. Seventeen volunteers aged from 18 to 25 took part in this experiment.
Significant effect of noise level and stimuli size on performance and reaction times was found. Performance in the most difficult condition (stimuli size 0.1 deg and noise level
40%) was 34%, which is a little bit higher than chance level (25%). Presentation of the congruent primer did not affect performance comparing to control condition (without primer or with incongruent primer), but it caused reduction of the reaction times in the most uncertain conditions (stimuli size 0.1 deg and noise level 40%).
Unconscious prime presented through magnocellular stream affect perception and decision making process on the object perceived by parvocellular pathway.
Supported by grant RFH No. 12-06-00947from the Russian government.
UDC 159.95 I. Vladimirov
IMPLICIT FORGETTING. WORKING MEMORY AS THE LOCUS OF PREVIOUS STORAGE SOLUTIONS
Ilya Vladimirov — Associate Professor at Yaroslavl State University, Yaroslavl, Russia, kein17@mail.ru
Research on mental set (Einstellung effect) is usually studied within the field of problem solving rather than memory research. However, the effect of mental set is related to memory effects. Once the subject is influenced by a mental set she/he tends not to see a simple solution but uses the habitual strategy that was formed in previous trials. We assume that previous solutions that form a set (Einstellung effect) are stored in working memory (WM), particularly — in a domain-specific part of working memory. We have used the water jar problem as a task (Luchins, 1942). In this study, we used between group plan with four conditions: three five-minute intervals between the setting and control trials that were used to perform another task and a control condition where there was no time span between the setting and the control trials. In the first of these three conditions inter-trial interval was filled with a talk with the experimenter, second — with solution of a puzzle (non-specific material), third — with arithmetic task (specific material). We compared the average time of problem solving in setting trials and control trials. For conditions 1, 2, 3 the differences are significant; the control trial is solved longer. Condition 4 doesn't show any significant differences between setting trial and control trial. Such results suggest the removal effect when the material of the task in the inter-trial interval interferes with the material of the main task. The study was not supposed to provide the complete verification of awareness of the principle solution. The possibility of awareness will be checked in subsequent experimental series.
This work is supported by Russian Foundation for Humanities (No. 12-36-01035a).
