This item is only available as the following downloads:
ANXIETY AND STATE DEPENDENCE IN PROCEDURAL MEMORY BY TARYA BARDWELL A Thesis Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the Sponsorship of Dr. Heidi Harley Sarasota, Florida May, 2013
ii Special Thanks To: Dr. Heidi Harley Dr. Gordon Bauer and Dr. Steven Graham Fellow Students Stephanie Altenbernd and Jonah Sant iago Friends and Family Whose invaluable help and support made this thesis possible. Acknowledgments
iii Acknowledgments.................................... ................................................... ........................ii Table of Contents.................................. ................................................... ...........................iii List of Figures.................................... ................................................... ...............................v Abstract........................................... ................................................... .................................vi Introduction....................................... ................................................... ...............................1 Implicit versus Explicit Memory for Movement....... ..............................................1 Motor Skill Acquisition............................ ................................................... .............5 Effects of Attention............................... ................................................... .................9 Cued Memory........................................ ................................................... ...............14 Emotion's Effects on Performance................... ................................................... ....20 The Current Study.................................. ................................................... ..............27 Method............................................. ................................................... ...............................29 Participants....................................... ................................................... ..................29 Materials.......................................... ................................................... ...................29 Procedure.......................................... ................................................... ..................29 Table of Contents
iv Results............................................ ................................................... .................................32 Acquisition........................................ ................................................... ..............32 Anxiety............................................ ................................................... ................34 Errors............................................. ................................................... .................36 Words per Minute................................... ................................................... ........37 Discussion......................................... ................................................... ..............................39 References......................................... ................................................... ..............................42 Appendix........................................... ................................................... ..............................46
v Words per Minute Across Acquisition Trials......... ................................................... ........32 STAI scores........................................ ................................................... ............................33 STAI Scores that Varied in the Expected Directions. ................................................... .....34 Errors............................................. ................................................... .................................35 Words per Minute................................... ................................................... ........................36 Words per Minute for Individuals with Ideal Scores. ................................................... ....37 List of Figures
vi ANXIETY AND STATE DEPENDENCE IN PROCEDURAL MEMORY ABSTRACT Tarya Bardwell New College of Florida 2013 Procedural memory is memory of a skill that is wel l-rehearsed and automatic. It is used during many common activities, such as playing spor ts, playing musical instruments, and typing. However, performance of these skills under pressure is often negatively affected by anxiety. Previous work on memory has determined that when one learns something in a specific state, one recalls that information best when in the same state. This statedependent memory is effective with a variety of sta tes, particularly with explicit memory, such as a list of words. How state-dependence appli es to procedural tasks is less well understood. The goal of this thesis was to apply s tate-dependent memory to procedural skills in hopes of reducing anxiety's negative effe cts on performance. To that end, skilled typists (40 total, 28 female, 12 male) practiced ty ping an unfamiliar passage in either an anxious or a relaxed state, and then typed the same passage again in either the same, e.g. anxious-anxious, or a different, e.g., relaxed-anxi ous, state. Participants' typing speed and errors did not differ across conditions, potentiall y because all participants relaxed over the course of the study. ______________________Dr. Heidi HarleyDepartment of Psychology
Running Head: ANXIETY AND STATE D EPENDENCE 1 Procedural memory is memory for a skill that is fam iliar and automatic. This type of memory is integral to many common activities, in cluding playing sports, playing musical instruments, and typing theses. Often, circ umstances demand that procedural tasks be executed while in an anxious state. Howeve r, anxiety reduces successful and efficient execution of a wide array of tasks, and i n some cases inhibition of these skills can be detrimental, e.g. driving in poor weather co nditions. The ability to perform a task effectively relies on 1) having learned it well in the past and 2) the a bility to retrieve it. While reducing or even elimi nating anxiety for affected individuals may be beneficial, it is not always feasible. On th e other hand, evidence indicates that people who learn and then recall information in the same emotional state will recall it better than those whose learning and recall states differ (Anderson, Einstein, and Ohalloran, 1989; Eich, Weingartner, Stillman & Gill in, 1975; Immink, Wright, and Barnes, 2012; McDaniel). This is called state-depen dent learning or simply state dependence. Perhaps state dependence, using the anx ious state in acquisition to one's advantage in hope of better recall, may provide a u seful approach to improving performance under difficult conditions. The current study will investigate how participants respond to learning and recalling in t he same anxious state. Implicit versus Explicit Memory for Movement Performances of a learned task rely on procedural m emory, which is implicit and typically results from amassed experience. However, the line between explicit and implicit memory can be blurry at times. In order to execute a task skillfully and efficiently, it needs to be explicitly understood, and then practiced until it is implicit. For
ANXIETY AND STA TE DEPENDENCE 2 example, Duke, Simmons, and Cash (2009) observed ad vanced piano majors in order to study influences on experts' acquisition of ideal t iming and fluidity. Participants were observed while they practiced a three-measure keybo ard passage until they reached target confidence. Participants were ranked for overall pe rformance quality. No significant relationship was found between retention tests and practice time, number of total practice performances, or number of practice trials. Three o f the 17 participants were especially skilled in their evenness of tone and fluidity of p laying, so researchers closely watched the practice techniques used by these three and com pared these techniques with the other participants. Strategies that these individuals us ed with greater frequency included playing with hands together and practice with infle ction early on when learning the new passage, thoughtful practice, and immediately addre ssing and attempting to fix errors. Apparently, practice strategies affected performanc e more than the amount of time spent practicing, potentially because the successful stra tegies supported the successful transition of explicit memory to implicit memory. D eliberate and conscious effort to incorporate tone and fluidity at an early stage mad e skillful piano playing automatic. Feedback also affects memory of movement (Lee & Mag ill, 1985). In this study, participants used a sliding apparatus; they were fi rst required to move the slide from a designated starting location to a target point on t he track, marked by a barrier. Following this, the barrier was removed and the participants' new task was to move the carriage to the place where the barrier had previously been. Th e researchers concluded that participants' movement representation was embedded with the memory; participants developed feedback reliance (hitting the barrier) w hen learning. Since the barrier became
ANXIETY AND STA TE DEPENDENCE 3 part of the memory for the movement, the movement b ecame challenging without the barrier. This is an example of how aspects not nece ssarily required for the movement itself assist in memory of the movement. This kind of feedback reliance can be found in our daily lives. If, when reaching for the light switch, one finds that is suddenly ab sent, this is likely to give rise to explicit awareness of the movement and possibly change futur e behavior. When typing, typists have an implicit expectation of where the keys are. If the keys are suddenly shuffled around, ones automatic motor movement no longer wo rks, and explicit knowledge must be used when re-learning the new system. Because pr ocedural motor memory results from familiarity with both individual movements and overall sequences, it can also be disrupted by changing the requirements of the seque nce (Gobel, Sanchez, & Reber, 2011; Shanks & Johnstone, 1999). Shanks and Johnstone (19 99) explored the explicit knowledge underlying a perceptual-motor skill by tr aining participants on a reaction time task using a keyboard. Participants responded to b oxes with letters in them, presented on a computer monitor. At least two preceding location s were needed to predict the next location of a box. Shanks and Johnstone (1999) mea sured reaction times and used these to show procedural memory. They found that a change in the sequence slowed down response times, indicating that participants needed more processing time to think about the task, making it no longer automatic. Here, with the occurrence of a change, the motor knowledge was no longer implicit. Manipulating a fa miliar task makes it novel, which changes whether or not knowledge is implicit or exp licit.
ANXIETY AND STA TE DEPENDENCE 4 Automatization of motor skills often requires learn ing sequences of precisely timed movements and perceptually cued tasks. Gobel, Sanchez, and Reber (2011) explored participants' acquisition of temporal and ordinal knowledge using a computer with a keyboard and a monitor. For training, four gold rings (targets) were displayed to the 63 participants, and their task to press the co rrect key when a cue was centered in its target across 1440 trials. Later, participants perf ormed an explicit recognition test, in which they were shown five 12-item sequences and ha d to identify whether or not they had practiced the sequence before. In the free reca ll test, the participants were told that there was a repeating sequence and that their job w as to reproduce the sequence. Participants made more mistakes during the pseudora ndom blocks than the repeating sequence blocks, which were presented immediately b efore and after the pseudorandom sequences, indicating that sequence-specific learni ng was used. The average recognition score indicated that the participants had some expl icit knowledge of the sequences. In addition, when the sequence or the timing requireme nt for the sequence was changed, performance returned nearly to novice level. This a lso indicates that the participants had once again to rely on explicit memory. Although the motor skill had already been developed, a change in the sequence required explic it awareness, requiring the participants, on some level, to relearn the skill. For example, when typing, an expert is likely not only to have an implicit knowledge of wh ere all the keys are and how to reach them, but also to have an implicit knowledge of ent ire words. These words can be considered procedural sequences; when confronted wi th a novel word the typist would
ANXIETY AND STA TE DEPENDENCE 5 have to slow down in order to explicitly consider w hat the letters are and how they are ordered. Hence, both explicit and implicit processe s occur in the same task. Studying an observers' knowledge of a sequence of m ovements may help tease apart implicit sequence knowledge and individual mo tor movements, since the observer does not become physically familiar with the proced ural motor movements involved. Dance is a form of nonverbal expression using compl ex phrases of movement (Opacic, Stevens, & Tillman 2009). Sequences may be consider ed grammatical or ungrammatical, depending on how conventionally th e moves are ordered. Sequences that recur and develop have a specific grammatical structure. An experienced observer may be able to identify a particular style based on this structure. Opacic, et al. (2009) studied how participants unfa miliar with dance could learn and recall the grammatical structures of these sequ ences in Australian contemporary dance. After participants in the exposure group obs erved grammatical dance, exposure and control groups watched and identified grammatic al dance structures in a test phase. The exposure group was more successful at this task with no correlation between confidence and accuracy, thereby suggesting that kn owledge was unconscious. Hence, these novice observers learned dance grammar throug h exposure, indicating an implicit learning effect, without actually executing the mov ements; through simple observation participants developed implicit, non-motor memory.Motor Skill Acquisition According to the direct matching hypothesis, a mech anism that maps an observed action onto motor representations is responsible fo r action understanding. Flanagan &
ANXIETY AND STA TE DEPENDENCE 6 Johansson (2003) hypothesized that patterns of eyehand coordination are similar when performing and observing a block stacking task. In this study, 16 right handed participants, unaware of the purposes of the study, worked in two groups. Eight participants both observed and performed the block stacking task. Each trial was followed by an unstacking trial. Eight other partic ipants observed the same task, but the demonstrator was hidden, controlling the blocks fro m below the work surface. Gaze position of the right eye and the position of the tip of the right index finger of the participant or actor were recorded with mini ature electromagnetic sensors. Flanagan and Johansson (2003) found that participan ts' gazes were predictive of an actor's hand, rather than reactive, in the block-st acking task. The participants' eye movements not only preceded the movement of the act or, but also were very similar to the way they moved themselves. These findings provi de strong evidence for the directmatching hypothesis and indicate that the way we th ink about motion is similar when both executing and observing movement. Procedural motor skill acquisition is a complex pro cess, affected by many elements, including the manner in which one practic es and the amount of time spent practicing (Duke, Simmons, Cash, 2009; Ericsson, Kr ampe, & Tesch-Rmer,1993), the amount and type of feedback received when practicin g (Lee and Magill, 1985; Park, Shea, & Wright, 2000), and individual differences i n anxiety sensitivity and attention control (Mesagno, Harvey, and Janelle (2012). Erics son, Krampe, and Tesch-Rmer (1993) conducted two studies on deliberate practice and its effects on performance. In the first study, 10 violin students at the Music Academ y of West Berlin were chosen for their
ANXIETY AND STA TE DEPENDENCE 7 potential as international soloists, and ten violin students were chosen from the music education department which had lower admission stan dards. Ten other violin students, who their professors called good violinists, also participated. Music education participants were assigned as music teachers. In addition to these groups, 10 middleaged violinists from two symphony orchestras in Ber lin were interviewed. Data collection procedures for the violinists were identical. First, biographical information was obtained, including start of practi ce, sequence of music teachers, participation in competitions, and number of hours per week spent practicing. Following this, participants were given instructions to respo nd to 10 categories of everyday activities and 12 categories of musical activities. They were asked to rate each on a scale from 0-10 in terms of relevance to improving perfor mance on the violin, effort needed to perform the activity, and enjoyment of the activity Participants also answered questions about practice and concentration, and recalled acti vities they did during the previous day. The researchers found that the best indicator of pe rformance was the success of performances at competitions. Also, the best violin ists reported a mean of 128.9 min of music they could play without preparation, and the good violinists averaged 79.1 min in this category. These averages were both higher than that of the music teachers, at 42.27 min. The participants who practiced with deliberati on were the best performers. For the second study, 12 expert and 12 amateur pian ists participated in this twosession experiment. The sessions were separated by a 7-12 day gap, during which time participants provided biographical information abou t their practice and experience. The task was to play a series of nine key-strokes eithe r with one or both hands on an electric
ANXIETY AND STA TE DEPENDENCE 8 keyboard. Each finger was assigned to a white key, so no lateral movement was necessary. Strings of nine numbers were presented, each indicating a finger. A display showed which numbers to play on a computer monitor, enabling the researchers to manipulate the complexity of hand coordination need ed. Participants were given two blocks of practice trials, with feedback on accurac y, speed, and timing during the second block. The only data used were from the third block which consisted of six trials. Near the end of the first session, participants began re cording daily activities for seven days between sessions one and two in a diary. After the debriefing for the diary procedure in Session 2, participants gave three performances of a piece by Bach. Then they completed the Digit-Symbol Substitution Test (DS), which was designed to measure perceptualmotor speed. A variety of factors were related to the participan ts' performance, but, once again, the most notable was deliberate practice. Ericsson, Krampe, and Tesh-Rmer (1993) pointed out that individual differences in performe rs are not because of innate talent, but rather intense and deliberate practice over a minim um of 10 years. Practice deliberation may result from responding pr operly to feedback during motor skill acquisition, playing a major role in th e success of automization (Lee & Magill, 1985; Park, Shea, &Wright, 2000). Feedback can be helpful, but too much guidance may also prove to interfere with making me mory implicit: the guidance hypothesis states that too much guidance can inhibi t learning. Park, Shea, and Wright (2000) contextualize the guidance hypothesis with a study that required participants to learn a static pressure task: Forty-eight undergrad uate students participated in this study.
ANXIETY AND STA TE DEPENDENCE 9 A device took static force measurements and transmi tted these measurements to a microcomputer. Participants lay on a table with a c omputer monitor above their eyes. The computer displayed a target waveform first, and par ticipants were instructed to apply the right amount of correctly timed pressure to match t he waveform. The first group received concurrent feedback during each trial and viewed an overlay of their performance on the target waveform at the end of each trial. This grou p was the 100% concurrent feedback/ 100% terminal feedback group (100%C100%T). The se cond group (50%C100%T) received concurrent feedback on only half of the ac quisition trials. In addition to these, there was the 0%C100%T and the 100%C0%T groups. T he conditions consisted of 10 blocks of 10 trials each. Park, Shea, and Wright fo und that the 50%C100%T and the 0%C 100%T did not significantly differ from each o ther, but both had fewer errors than the 100%C100%T and the 100%C0%T groups. Strong gu idance effects were found for all the 100%C conditions, indicating that concurren t feedback actually reduced retention of a newly learned motor skill.Effects of Attention Attention affects performances among both experts ( those relying more on procedural memory) and novices (those relying more on explicit memory). For example, two experiments by Beilock, Carr, MacMahon, and Sta rkes (2002) studied the relationship between attention and sensory skills. In their first experiment, undergraduate students with two or more years of golf experience were instructed to putt a golf ball with as much accuracy as they could from nine different locations. In the skill-focus condition, golfers were instructed to focus on the moment they finished the follow-through of their
ANXIETY AND STA TE DEPENDENCE 10 swing, and to say stop and stop the club upon com pletion. In the dual-task condition, participants were instructed to listen to a series of tones, and say tone every time they heard a target tone and, at the same time, putt the golf ball. All participants experienced both the skill-focus and dual-task conditions, perf orming 20 putts in each. Putting accuracy was measured by the distance between the b all's stopping point and the target. Experienced golfers monitoring their putting had th e capacity to pay attention to a secondary task, such as a tone. When instructed to attend to the stop at the end of the golf swing, however, even experienced golfers' performan ce decreased. Beilock et al. (2002) argued that the focus of attention directed on a sp ecific movement may account for choking under pressure. Experiment 2 continued the investigation of attenti on and motor tasks, this time with dribbling a soccer ball. Participants were 10 novices and 10 experts and were instructed to dribble with either their dominant or non-dominant foot. Their task was to dribble a soccer ball as quickly as they could thro ugh a slalom course. In the skill-focused condition, the participants heard a single tone and a target word. They were instructed to say which side of their foot was contacting the bal l when they heard the tone. In the dualtask condition, they were required to listen to the tape of random words and repeat them out loud whenever they heard the target word, thorn When dribbling with the dominant foot, experts perf ormed best in the dual-task condition, while novices performed best in the skil l-focused condition. When dribbling with the non-dominant foot, all participants perfor med best in the skill-focused condition. When executing a task that is not yet procedural, r esearch indicates that optimal
ANXIETY AND STA TE DEPENDENCE 11 performance is achieved by focusing on individual m ovements. Dribbling with the nondominant foot is likely non-procedural memory for e veryone, but experts probably had dominant-foot dribbling in procedural memory. These results support previous findings about attention and skill acquisition. In order to achieve optimal performance, a novice focuses on individual movements in order to learn t hem, while an expert focuses on goaloriented aspects. Based on the research of Beilock, Carr, MacMahon, & Starkes (2002), researchers Ford, Hodges, & Williams (2005) conducted a study t o compare performance of highand low-skill soccer players across different level s of attentional focus. Participants were 10 skilled and 10 non-skilled soccer players. Like the participants in Beilock et al.'s study, their task was to dribble a soccer ball arou nd a slalom course. They were directed to do the task quickly and accurately. There were three conditions. In the control condition, no attentional directions were given. In the external focus condition, participants were to give an indication every time they heard a single tone, randomly occurring on tapes, played every 6 s, or to repeat a word out loud. In the internal focus condition, participants paid attention to their fee t and arms. For the internal skill-relevant condition, they were instructed to indicate which s ide of their foot was contacting the ball; for the skill-irrelevant condition, they were asked to say which arm was nearest to the next pylon. The dependent measures were the tim e taken to complete a single errorfree trial, as well as the number of errors made du ring other trials. Ford et al. found that internal foci generally reduced performance. The sk illed performers were more affected by these tasks than the less-skilled performers. No vices were negatively affected by the
ANXIETY AND STA TE DEPENDENCE 12 arm (irrelevant focus) and the skilled performers w ith more implicit knowledge were negatively affected by both the arm and foot. An experiment by Duke, Cash, and Allen (2011) sough t to apply knowledge about external focus to musicians who experienced auditor y feedback during their performances. Sixteen music students (8 males) part icipated in this study. All participants experienced all conditions. Prior to each session, participants were given verbal and written instructions to direct their attention to e ither the keys, the piano hammers, or the sound, depending on the condition. Each of the four conditions started with four training sessions, followed by a five minute delay, then a r etest, then a two minute delay, and then a final test. Duke et al. collected data on timing and volume by measuring the intervals between tones and the downward key velocities. Timi ng affected focusing on the goal, but volume did not. Focus on the sound led particip ants to play with far more even timing than those who focused on their fingers. Performanc e differences between other foci were not statistically significant. McNevin, Shea, and Wulf (2003) further explored the effects of attention on motor skills by testing how distance of an external focus affects performance. Participants balanced on a stabilometer, a platform with a fulcrum in the middle, and were asked to place each foot on either end of the platform and keep it as horizontal as possible. There were markers on the platform indica ting where the feet were supposed to go, and markers in front of the feet, farther out a nd on either side of the feet, and in between the feet. These markers were used for three experimental external focus study conditions, in which the participants were instruct ed to focus on one of the pairs of
ANXIETY AND STA TE DEPENDENCE 13 markers. The internal focus group was instructed to keep their attention on their feet. In order to control for visual effects, participants w ere asked to look straight ahead at the wall during the task, but mentally focus on the loc ations designated for their group. Participants were given two days of practice, durin g which they were reminded of where their focus was supposed to be. On the third day, t here were seven retention trials without instructions or reminders. The data were collected by measuring the boards' de grees from horizontal. The farther distances of the foci, the more accurate th e performance. Focusing on close effects and focusing on the marks closest to the feet led t o the poorest performance. Because this study was designed with practice time, McNevin, She a, and, Wulf concluded that focusing on something more distal actually improved not only immediate performance, but also learning of a motor skill. Some people's performance suffers when doing a task under pressure. People will do more poorly at a task when under pressure than t hey would otherwise. Possible reasons for this include distraction, i.e., focusin g on the anxiety more than the task at hand, and explicit monitoring, i.e., focusing too m uch on the mechanics of the task. DeCaro, Albert, Thomas, and Beilock (2011) conducte d four experiments to determine which, if either, of these theories explains perfor mance for different kinds of tasks. In experiment one, participants completed a rule-based or information-integration categorylearning task in a single-task condition, followed by one of two types of secondary task conditions. One of these secondary tasks was a dist racting task designed to divert attention, and the second was an explicit monitorin g task intended to prompt attention to
ANXIETY AND STA TE DEPENDENCE 14 the step-by-step components of learning. If categor y learning relies on explicit memory, it should take longer to learn that type of structure when simultaneously performing a distracting task than an explicit monitoring task. Roughly half the participants were randomly assigned to the distraction secondary task and the rest were assigned to the explicit monitoring secondary task. Participants completed a category-learning task ind ividually on the computer. This task required them to place each stimulus into a category (A or B) by pressing one of two keys. After each response, the computer disp layed whether or not the participant was correct. Participants' performance was measured until they successfully completed eight trials in a row. DeCaro, Albert, Thomas, and Beilock found that rule-based learning led to more errors when individuals were distracted Rule-based learning was unaffected, however, in the explicit monitoring condition. To s ee how performance was affected by anxiety, the experimenters added either low or high performance pressure for experiment two. Participants in the low-pressure conditions ex perienced no differences in performance, but those in the high-pressure conditi ons struggled with rule-based learning. In short, explicit monitoring and distraction hurt performance on different tasks. Cued Memory One's memory of something usually relies on some cu e, whether it is some external context cue or an internal state of mind. These cueing effects have been studied by many researchers. For example, Miles and Hardman (1998) studied memory dependence on physiological state by asking partici pants to memorize lists of words, either while exercising or while not exercising. Th ey found that participants remembered
ANXIETY AND STA TE DEPENDENCE 15 a list of words better while doing aerobic exercise if they also learned them while doing aerobic exercise. In fact, participants who learned and recalled while exercising performed just as well as those who learned and rec alled at rest. Lang, Craske, Brown, and Ghaneian (2001) conducted a similar study of state dependence, this time manipulating fear. They first selected for participants who had spider or snake phobias. These participants learned a list of words, after either being with their feared animal (caged), or in a recliner and l istened to calming music. The participants were either relaxed or aroused again p rior to the recall phase, in which participants had to recall as many words as they co uld. Participants who learned the words in a fearful state recalled them more success fully in while in a fearful state versus those who learned or recalled in different states. Eich, Weingartner, Stillman & Gillin (1975) studied state-dependent memory of words and word categories using marijuana. Particip ants memorized word-lists after administration of marijuana or a placebo. Participa nts engaged in either free-recall or cued-recall after either marijuana or placebo readm inistration. Eich, et al. (1975) found that recall of words was state dependent, even when encoding during the drug dissociated state induced by marijuana. Also, participants in c ued recall conditions performed better than participants in free recall conditions. This r esearch provides evidence that, even with states that generally decrease semantic memory (mar ijuana), this memory can be enhanced by inducing the state for both encoding an d recall. This example is parallel to the hypothesis that anxiety, which reduces motor sk ill performance (Mesagno, et al., 2012; Beilock, Carr, MacMahon, & Starkes, 2002), ma y actually improve motor memory
ANXIETY AND STA TE DEPENDENCE 16 when anxiety is applied in a state-dependent contex t. If this is true for semantic memory and dissociation, it could be true for motor memory and anxiety as well. State can affect motor memory. A study by Immink, W right, and Barnes (2012) examined the relationship between cutaneous tempera ture as a contextual condition and motor skill acquisition. Participants were randomly assigned to one of two training temperature conditions (hot or cold) and then were instructed to lie down, facing a monitor that was mounted above them. The left forea rm was placed inside the water sleeve which circulated either hot or cold water. F or the hot water, the water was approximately 40o C for the hot condition and approximately 15o C for the cold. Each of the participants practiced a force production hand grip task for 100 trials. After participants rested for 40 minutes, they were rando mly assigned to either a hot or cold training condition and a hot or cold test condition This resulted in four training-test conditions: hot-hot, cold-cold, hot-cold, and coldhot. Performance was based on the specification of force and error variability. As pr edicted, temperature changes reduced participants' accuracy, particularly after hot trai ning conditions. There are numerous studies confirming state depende nce having an effect in a variety of recall tasks. The majority of these test explicit memory, particularly memory of word lists. Memory cues, however, do not always enh ance recall, and looking at these simple word-list studies may give an over-simplisti c impression of cued memory. For example, a study by McDaniel, Anderson, Einstein, a nd Ohalloran (1989) did not find context effects across the board. These researchers studied place dependence, a.k.a environmental contexts that serve to cue memory. Tw o different rooms were used, one
ANXIETY AND STA TE DEPENDENCE 17 was full of caged rats and was not. Participants we re asked to recall twenty-four sentences, half of which were bizarre and half of w hich were not. Participants saw first a sentence, visualized it, and then rated its vividne ss on a rating from 1 ( clear, vivid image ) to 5 ( not very clear image ). In the first experiment the participants were requi red to solve division and multiplication problems after listening and respond ing to the sentences. Participants then moved to their designated recall environments (eith er the same or different) and then wrote as many of the sentences as they could rememb er. McDaniel, Anderson, Einstein, and Ohalloran found that there were no advantages t o returning participants to the original environment. In the second experiment, how ever, participants were asked to rate sentence typicality rather than visualize the sente nces. This time, a context-dependency effect occurred. The procedures for a third experiment were similar to those in the previous experiments. The participants in the imagery encod ing conditions were instructed to give vividness ratings, and the participants in the noni magery encoding conditions were instructed to give typicality ratings. Reinstatemen t of context significantly helped participants recall the sentences and, unexpectedly this context-dependency had a stronger effect with the imagery group than the noimagery group. This contradicts findings in previous experiments as well as pilot s tudies used by these researchers. Experiment 4 was designed to further examine the co ncept that the encoding task may determine the degree to which context reinstate ment influences recall. Half of the participants were given nonimaginal instructions an d half were instructed to group the
ANXIETY AND STA TE DEPENDENCE 18 sentences by finding some means of relating them to each other. In the grouped-sentence encoding condition, no context-dependence effects w ere found, but in the isolated sentence condition, recall was affected by context. Recall of common sentences in the same-environment condition was better than in then in the different environment condition, but the same context-dependency was not apparent for bizarre sentences. A fifth experiment took into account the effects of self-reference encoding, with the hypothesis that context-dependency would not be apparent in self-reference encoding. The context rooms in this experiment consisted of a sparsely furnished laboratory room and a much larger luxurious conference room. Partic ipants in the self-reference processing condition had to judge whether or not ea ch phrase described a specific event or object that they had experienced. Following this the participants were led to the other room or returned to the same room, depending on the condition. Then they were given a 10 min free-recall test. In the self-reference task, encoding tended to be m ore successful when the context was different, than when it was the same. Based on all of their experiments, the researchers proposed that encoding activities that promote retrieval routes lead to a reduced reliance on contextual cues and the effecti veness of context-dependency. Context-dependency is also affected by task difficu lty. Wright and Shea (1991) seated participants in front of a computer monitor and keyboard. The participants responded to a display on the screen indicating whi ch keys to press. The display provided both incidental and intentional stimuli. The letter s displayed were intentional, but they also varied in color, shape of key designations, ac companying tone, and position on the
ANXIETY AND STA TE DEPENDENCE 19 screen. The participants used intentional stimuli t o execute sequences when the incidental stimuli were not mentioned. The participants' reten tion was measured in three different conditions. In the same intentional-switched incide ntal condition, context from acquisition was reinstated. In the same intentional -switched incidental condition, intentional stimuli were presented with different i ncidental stimuli. Lastly, the no intentional/same incidental condition, incidental s timuli were presented without intentional stimuli. Participants in the same inten tional/same incidental retention condition performed better (with less error) than t hose who tested in the intentional/switched incidental retention condition Participants performed poorly when incidental stimuli were removed or changed between acquisition and recall. These results are an indication of context dependency for acquisi tion and recall, even if the participants may not be fully aware of their use of the incident al stimuli. Experiment two tested the hypothesis that dependenc y on context decreases with easier motor tasks. The procedures of the first exp eriment were repeated, but requiring only three key strokes instead of four. In this cas e, switching or removing the incidental stimuli made no difference on retention performance There was also a reduction in task difficulty. Because contextual differences did not affect performance in this experiment (unlike experiment 1), it can be concluded that eas ier tasks are less reliant on contextual cues. Experiment three was developed not only to further test this hypothesis, but also to confirm reliance on contextual stimuli. The find ings supported the results of the previous two experiments. Context dependency was moderated by difficulty. In addition,
ANXIETY AND STA TE DEPENDENCE 20 contextual memory decreased as the retention delay increased, such that the contextual stimuli became negligible in minutes. A later study conducted by Wright and Shea (1991) o n the effects of delayed versus immediate retention found that differing con textual stimuli and long delay between acquisition and recall negatively affected the production of the correct typing sequence either because contextual information was inaccessible after some time, or the source of information was not actively reinstated a t the time of the test. In another followup, half the participants experienced nine exposure s of the original context before the test and had fewer errors for the same retention context versus the different retention context. State dependent memory is a nuanced effect. Emotion's Effects on Performance Emotion is a state that affects memory and performa nce both as a cue and because movement itself can be related to mood. For example, Cassanto and Dijkstra (2010) found a relationship between emotion and the direction of movement. In an experiment that involved moving marbles up or down with positive or negative memories, they found that participants moved the ma rbles fastest when upward movements were paired with positive memories and do wnward movements were paired with negative memories. Emotion also has a direct effect on memory. Robinso n & Compton (2006) conducted two studies examining the priming effects on lateral spatial attention. Specifically, they studied emotional significance ( valence) and arousing priming effects on the symmetry of performance on a spatial memory task. In this study, two projectors
ANXIETY AND STA TE DEPENDENCE 21 were placed behind a participant's chair and were a ngled downward, so that the participant could easily view the images. One proje ctor displayed emotional slides and the other displayed dot slides. There were negative high-arousal slides, negative lowarousal slides, positive high-arousal slides, and positive low-arousal slides. Participants were instructed to indicate the number of dots they saw by pressing buttons. Participants were told that they were supposed to pay attention to the slides but were not informed that the valence of the slides was relevant. Dot discriminations were more accurate following po sitive slides than following negative slides. This evidence indicates that negat ive stimuli may interfere with a memory task because they are emotionally distractin g. Also, responses following positive slides were faster than responses following negativ e slides. Low arousal slides led to faster performance for both visual field targets, b ut for high-arousal slides there was faster performance for left visual field targets. Anxiety also affects memory because exposure to str ess can impair or enhance memory, attention, and executive functions. Areas o f the nervous system known to react to stress are the hypothalamic-pituitary-adrenal ax is ( HPA) and the sympathetic nervous system (SNS) (Hidalgo, Villada, Almela, Espin, Gome z-Amor, and Salvador (2012). By studying the reactivity of the HPA and SNS to psych osocial stress, Hidalgo et al. also studied the effects this stress has on short-term i mplicit memory and declarative memory. The enzyme alpha amylase and the hormone cortisol a re both linked to stress. Hidaglo et al. sampled these in order to test stres s manipulations. Fifty-two participants
ANXIETY AND STA TE DEPENDENCE 22 provided the initial cortisol saliva sample after f ive minutes in the lab. Then they received further instructions followed by the first salivary alpha-amylase (sAA) sample. In the stress condition, participants were subjecte d to the Trier Social Stress Test (TSST). The stress task occurred in the presence of a two person committee, a video camera, and a microphone. The task was 5 min of fre e speech (job interview), an sAA sample, a 5-min arithmetic task, then another sAA s ample. Following 20 mins. recovery, cortisol and sAA samples were taken again. Then the participants performed two memory tests: one for non-declarative memory and one for d eclarative memory. The participants repeated six trials, took a 30-min. break (with a c ortisol test during the break), and then took two more memory tests. In the control conditio n, participants completed a task similar in workload to the stress task, but without the stress provocation. They read aloud for 5 min and counted for 5 min. The timing of the saliva samples was the same in both conditions. Cortisol levels increased immediately a fter the stress task and decreased during recovery, while sAA increased continually un til the end of speech. Cortisol and sAA reactivity did not correlate with declarative m emory at all. However, implicit memory actually improved. McCabe (1999) explored the relation between anxiety sensitivity and memory for threat words. The Anxiety Sensitivity Index (ASI; R eiss et al., Peterson & Reiss, 1992) identified undergraduate participants with scores h alf a standard deviation above or below the norm as high anxiety sensitivity (HAS) an d low anxiety sensitive (LAS). There were 38 HAS participants and 36 LAS participants. A rousal, anxiety, and word type were each studied in relationship to each other. Each pa rticipant experienced the four word
ANXIETY AND STA TE DEPENDENCE 23 types, but the arousal and anxiety manipulations va ried between participants. A word list of 24 neutral words, 24 positive words, 24 threat w ords, and 24 anxiety words was used. McCabe (1999) used controlled breathing to manipula te arousal. There were low and high arousal groups; participants in the high-arous al condition listened to a tape instructing them to breathe at a rate of 30 breaths /min for 5 min., after which they saw and rated how much they liked the words. Participan ts performed a word stem completion test to assess implicit memory and a cue d recall test to assess explicit memory. Following these tests, participants took th e Anxiety Sensitivity Index, StateTrait Anxiety Inventory, Beck Depression Inventory, Emotionality Rating Form, and Visual Analogue Scales. No group differences in word completion data were f ound in implicit memory for the four information types, but there were group di fferences in explicit memory. High anxiety participants explicitly remembered threat w ords better than positive, neutral, or anxiety words. Low anxiety sensitivity participants had better explicit memory of positive words, as compared to neutral, threat, or anxiety w ords. The link between anxiety and threat words suggests that, in certain situations, stress may assist explicit memory. However, these findings are limited by awareness, s ince implicit memory was unaffected by threat words for both HAS and LAS participants. In other research, Bond and Omar (1990) studied sta te-dependent learning by looking at a specific recurring problem for perform ers, called the next-in-line effect. The next-in-line effect is to the inability to recall w hat happened just before a performance. This effect may be caused either by a deficit in en coding (storing information into
ANXIETY AND STA TE DEPENDENCE 24 memory) or recall (accessing the memory later). Bon d and Omar proposed that problems with recall specifically could be caused by state-d ependence, and that the solution would be re-inducing the state that participants were in before performance. In this study, 95 participants were divided into 4 groups. In each group, twentyfour numbered desks were arranged in a circle with a participant in each. A stack of index cards was face down on each desk, with a word on th e bottom of each card. A tape recording played a number from one to twenty-four, every five seconds. In response to the number, the participant at that desk read the t op index card aloud. Participants were told that this was a memory test and were instructe d to pay attention to the words. Half the participants read aloud and half just listened, manipulated by the tape recording calling only odd or only even numbers. Then the par ticipants were given three minutes to write down as many words as they could remember. Pa rticipants were also divided into induction and re induction groups. There were four experimental trials, each with thre e phases: induction, reinduction, and validation. For the induction phas e, half the participants read and half listened; and for the reinduction phase, the reader s read a different set of words. All participants had to write the words from the induct ion phase while engaging in the reinduction phase (they had to pause their writing if their number was called). For the validation phase, participants wrote words from the reinduction phase in silence. A state that was induced when participants are next -in-line in the first phase would be reinduced when next-in-line for the second phase. Validation of the next-in-line effect takes place in the third phase. The reader c onditions tested recall ability for those
ANXIETY AND STA TE DEPENDENCE 25 who and to both read and listened and the listener condition tested recall ability of participants who only listened to the words, but di d not read them out loud. Following these trials, participants filled out a social anxi ety scale. Bond and Omar found that participants remembered mo re words from the induction phase than the reinduction phase, which m ay be explained by the fact that participants' were attempting to retrieve a set of words and encode a separate set of words at the same time, straining their attentional capac ities. Participants with high social anxiety were found to suffer from the next-in-line effect more than those who had low social anxiety. Bond and Omar proposed that the next-in-line effect is a state-dependent retrieval deficit, meaning that people struggle more to recal l when in a relaxed state if they learned in an anxious one. By examining their roles (reader or listener) and the amount of social anxiety that was present during the reinduction pha se, they could test their hypothesis. Since, after the first phase, half the participants took turns reading aloud and the other half recalled the first set of words, Bond and Omar assumed that their participants would be in the same state before their second performanc e that they experienced for their first. Comparing recall at the reinduction phase, Bond and Omar again found a strong next-inline effect for high anxiety individuals. If their state-dependence hypothesis was correct, however, the next-in-line effect would be reduced o r eliminated for participants at reinduction of the anxious state. This was not the case. Participants experiencing high social anxiety did not recall what happened prior t o performance even when anticipating another performance at the time of recall.
ANXIETY AND STA TE DEPENDENCE 26 Bond and Omar provided strong evidence against stat e-dependent memory affecting performance, at least for the next-in-lin e effect. This effect, however, was measured in terms of explicit, non-motor memory. Ho w procedural memory is affected by anxiety in terms of state-dependence remains to be studied. McCabe (1999) and Bond and Omar (1990) both provide d evidence that explicit memory varies between individuals based on their se nsitivity to anxiety. Mesagno, Harvey, and Janelle (2012) looked specifically at t hese differences in relation to procedural memory in a study of choking in skilled basketball player. Basketball players were chosen if they had a minimum of 5 years in a c ompetitive basketball league and extreme scores on either end of the brief fear of n egative evaluationII (BFNE-II). Standard basketball equipment and facilities were u sed, with five separate shooting areas on a basketball court. Participant measures include d age, gender, experience, highest level of competitive basketball played, and state a nxiety via the Revised Competitive State Anxiety Inventory-2 (CSAI-2R). During the bas ketball shooting task, participants' accuracy and preparation time (the interval between when the participant received the ball to the time it was released) were measured. Th ey were each allowed 50 shot attempts. Each participant took part in three phase s (familiarization, low-pressure, and high-pressure). For the familiarization phase, only the researcher was present. After taking the CSAI-R2, 10 warm-up shots and the 50 regular shots were performed. The order of shooting areas was randomized to reduce the chances of fatigue or practice inaccurately skewing the results. The low pressure phase was ide ntical to the familiarization phase,
ANXIETY AND STA TE DEPENDENCE 27 except for being called a research experiment whic h was shown to make participants more nervous. For the high-pressure phase each participant receiv ed specific instructions designed to induce anxiety, by adding monetary ince ntives, a video camera, and a live audience. Monetary incentives increased with the nu mber of successful shots. Participants were falsely told that video recording s were to be used for bio-mechanical analysis. Audience members were the research assist ant operating the camera and four sports science students. With the exception of the pressure manipulation, the highpressure phase was identical to the familiarization and low-pressure phases. Based on the FNE scores, Mesagno et al. (2012) dete rmined that the anxiety manipulations were effective. They found no signif icant change in performance between lowand highpressure phases for the low-FNE grou p, but a significant decrease in performance between lowand highpressure phases for the high-FNE group. Hence, choking results when people who fear negative evalu ation are asked to perform under pressure.The Current Study Anxiety can have a detrimental effect on the perfor mance of a skill (Mesagno, Harvey, & Janelle, 2012). This decrement in perfor mance could be due to a disruption of motor memory skills in a new state, i.e., an anxiou s state. Declarative memory is better when people recall information in the same state in which they acquired it (e.g., Eich, Weingartner, Stillman & Gillin,1975; Lang, Craske, Brown, & Ghaneian, 2001; Miles & Hardman,1998). Similarly, procedural memory was fac ilitated when people with cold or
ANXIETY AND STA TE DEPENDENCE 28 hot skin temperatures reenacted a skill they learne d when they were experiencing the same skin temperature (i.e., cold-cold, hot-hot) (I mmink, Wright, & Barnes, 2012). Might practicing a skill in an anxious state reduce the l ikelihood of a performance decrement when that skill is reenacted later in an anxious st ate? The current study will address this question by having skilled typists rehearse typing a passage in either an anxious or relaxed state after which they will reproduce that passage again in either the same state (e.g., anxious acquisitionanxious recall) or a di fferent state (e.g., relaxed acquisitionanxious recall).
29 Method Participants Participants were 40 (12 male, 28 female, mean age = 20.45) students at a liberal arts college in Southwest Florida who volunteered f or the study by responding to an advertisement on an online forum or by talking dire ctly to the researcher. Because procedural memory was the target interest, particip ants were selected for their typing skills. To qualify for participation, they had to b e able to type with all ten fingers at a minimum of 35 words per minute (WPM).Materials Participants filled out the State section of the St ate-Trait Anxiety Index (STAI) and took a short survey about typing experience. Se e Appendix for these sources. Video recordings of another person and of an ant-lion dig ging a hole were also presented. A 291-word excerpt from This Side of Paradise (F. Scott Fitzgerald), taken from the Project Gutenburg website was used for typing acquisition a nd recall. The excerpt was selected for its length; no participant was expected to type fast enough to complete it (145 WPM) All activities were completed on a laptop computer. Procedure In order to test the effects of state on procedural memory, anxiety was manipulated for two phases of the study: acquisitio n and recall. There were four groups total: relaxed acquisition-relaxed recall, relaxed acquisition-anxious recall, anxious acquisition-relaxed recall, and anxious acquisition -anxious recall.
30 Participants were first told that the experiment wa s designed to study typing skills in college students and were asked to sign an infor med consent form. Then a participant was assigned either to the anxiety or the relaxed c ondition for the acquisition phase. In the anxiety condition, a participant saw a video o f the a person watching a monitor and was told that the watcher was evaluating the partic ipant's attractiveness. A participant in the relaxed condition was told that the watcher was simply another participant that the experimenter was monitoring. All participants saw the same fake webcam video of the same person in all conditions. In the acquisition phase, participants typed the F. Scott Fitzgerald excerpt five times, for two minutes each time. Participants expe rienced either the anxious or the relaxed manipulation during this phase. Then they f illed out the state part of the STAI. Following this, participants took a typing experien ce survey and watched the video of the ant-lion in order to allow time for the mood manipu lation to wear off. Following the ant-lion video, participants were ag ain assigned to the anxiety or relaxed condition for the recall phasethe conditi ons were identical to those in the acquisition phase, i.e., the person on camera was a n evaluator (anxiety condition) or another participant (relaxed condition). Ultimately an equal number of participants were assigned to each acquisition-recall group (i.e., an xiety acquisition-anxiety recall; anxiety acquisition-relaxed recall, relaxed acquisition-rel axed recall, and relaxed acquisitionanxious recall). Participants typed the excerpt one last time during this manipulation, again with a two minute time limit.
31 Upon completion of the study, the purposes of the a nxiety manipulation were explained in detail, and the participants were than ked for their time.
32 Results Data from 4 of the 40 participants were omitted: on e of these participant's typing speed was below the minimum (35 WPM) and three part icipants failed to answer all the STAI questions. Therefore, data from a total of 36 participants were used for analysis: 8 in the relaxed-relaxed group, 9 in the anxious-anxi ous group, 10 in the anxious-relaxed group, and 8 in the relaxed-anxious group.Acquisition A practice effect occurred across the five acquisit ion trials: trial 1 (M=51.93 WPM, SD = 11.00), trial 2 (M=54.43 WPM, SD=10.67), trial 3 (56.65 WPM, SD=14.60), trial 4 (M= 55.87 WPM, SD=10.90), and tr ial 5 (M=57.18 WPM, SD=11.68). Trial 1Trial 2Trial 3Trial 4Trial 5 0 10 20 30 40 50 60 70Words per Minute across Acquisition Trials
33 However, as WPM increased, so did errors, indicatin g that as participants typed faster they made more mistakes: trial 1 (M=1.94 err ors, SD = 1.50), trial 2 (M=1.83 errors, SD=2.46), trial 3 (2.3 errors, SD=3.58), tr ial 4 (M= 2.58 errors, SD=4.18), and trial 5 (M=2.78 errors, SD=3.96). Trial 1Trial 2Trial 3Trial 4Trial 5 0 0.5 1 1.5 2 2.5 3Errors across Acquisition Trials
34 Anxiety The anxiety manipulation did not work. An ANOVA was conducted on the STAI scores; the scores for the acquisition phase were n ot significantly higher in the anxiety manipulation (M=44.95, SD=11.07) than the relaxed c ondition (M=42.98, SD=10.31), t(16)=0.53, p=0.60. Similarly, STAI scores for the recall phase were not significantly higher in the anxiety condition (M= 38.72, SD=9.50) than the relaxed condition (M= 37.33, SD=6.62), t(17) = 0.47, p=0.64. Relaxed-Relaxed Anxious-Anxious Relaxed-Anxious Anxious-Relaxed 0 5 10 15 20 25 30 35 40 45 50Mean STAI Scores across Groups Acquisition STAIRecall STAI
35 Twelve participants had STAI scores that varied in the expected directions: 2 in the relaxed-relaxed group, 2 in the anxious-anxious group, 1 in the relaxed-anxious group, and 7 in the anxious-relaxed group. Relaxed-Relaxed Relaxed-Relaxed Anxious-Anxious Anxious-Anxious Relaxed-Anxious Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed 010203040506070STAI Scores that Varied in the Expected Directions Acquisition WPMRecall WPM
36 Errors An ANOVA was conducted on recall errors; average re call errors did not vary across the four conditions, F (3) = 2.105, p = 0.1 19: relaxed(M= 2.0, SD=1.31)-relaxed (M=2, SD=2.3); anxious(M=4.11, SD=3.73)-anxious (M= 7.11 SD= 5.84); relaxed(M=1.67, SD=1.87)-anxious (M= 1.33, SD=1.73) ; anxious (M=3.20, SD=6.36)relaxed (M=3.50, SD=7.96). Because there were no si gnificant differences, no post-hoc tests were conducted. A GLM procedure was also cond ucted on errors; no significant differences were found F(1)=1.94, p=.17. Relaxed-RelaxedAnxious-AnxiousRelaxed-AnxiousAnxiou s-Relaxed 0 1 2 3 4 5 6 7 8Average Errors across Group AcquisitionRecall
37 Words Per Minute An ANOVA was conducted on recall WPM; average recal l WPM did not vary across the four conditions, F (3) = 0.75, p = 0.53: relaxed(M=57.88, SD=16.30)-relaxed (M=57.40, SD=15.44) ; anxious(M=59.0, SD=7.75)-anxi ous (M=61.56 SD=10.09); relaxed(M=50.06, SD=10.50)-anxious (M=54.06, SD=9.9 9) anxious(M=60.75, SD=10.03)-relaxed (M=56.15, SD=8.17). Because were no significant differences, no post-hoc test were conducted. A GLM procedure was also conducted on WPM; no significant difference was found, F(1 )=.03, p=.87. Relaxed-Relaxed Anxious-Anxious Relaxed-Anxious Anxious-Relaxed 0 10 20 30 40 50 60 70Mean Words Per Minute across Group AcquisitionRecall
38 The WPM for participants who had predicted STAI sco res showed that three of the four participants who were in the same conditio ns for both acquisition and recall (relaxed-relaxed and anxious-anxious) had the same WPM for both acquisition and recall. Four of the eight participants who were in differin g conditions between acquisition and recall performed better during recall than acquisit ion. Relaxed-Relaxed Relaxed-Relaxed Anxious-Anxious Anxious-Anxious Relaxed-Anxious Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed Anxious-Relaxed 01020304050607080Words Per Minute for Individuals with Expected STAI Scores Acquisition WPMRecall WPM
39 Discussion The purpose of this study was to explore possible s tate-dependent effects of anxiety on procedural memory. Therefore, capable ty pists practiced a prose passage when they were anxious or when they were relaxed, and th en they typed that passage again either in the same or different conditions. Howeve r, the four groups did not change differentially across acquisition and recall. In pa rt this may have been because everyone became more relaxed over the course of the experime nt. People did not appear to be especially anxious at being evaluated. Perhaps they did not believe the cover story, or perhaps they became too focused on the typing task to be anxious about the evaluation. Some participants casually mentioned that they did not believe that the webcam manipulation was live, or that they were not concer ned about attractiveness ratings. Being timed and under pressure to type quickly may have a ffected anxiety more than the anxiety manipulation. Mesagno, et al. (2012) used multiple techniques to manipulate anxiety, including the use of a video camera and a live audience. They informed participants that their basketball performance was being analyzed on the vi deo recordings; a manipulation that directly related to the task at hand (unlike attrac tiveness and typing). Similarly, Hidalgo et al. (2012) used a live audience (a panel of two) a video camera, and a microphone in an interview situation to manipulate anxiety. Their alpha amalyse (sAA) and cortisol samples indicated that these anxiety manipulations were successful. Participants' doubt of the webcam in the current experiment may have be en eliminated by using a live audience; participants would have known for sure th at they were being watched. In
40 retrospect, perhaps the experimenter could have bee n the analyzer by changing whether or not I looked over the participants' shoulders du ring typing or did something else during their typing. Another technique would be to select for participan ts who are naturally more prone to anxiety. Mesagno et al. (2002) did this by selecting for participants with highFNE scoresthese participants were more responsive to the anxiety manipulations and choked more than the low-FNE participants. High-FNE participants would make ideal participants not only due to their proneness to anx iety, but also because they are the people who could benefit most from state-dependent learning if it was found to work. Participants who respond strongly to anxiety manipu lations, whether due to stronger manipulations or natural anxiety proneness would probably need longer recovery times from the manipulation than the anxie ty manipulation applied here. For example, Hildalgo et al. (2012) used a 20 minute re covery time after the anxiety manipulation, and measured a decrease in cortisol, indicating that a minimum of 20 min. should have passed between acquisition and recall. Skill acquisition is difficult to study because it can take so long. Typing was an accessible skill, and acquisition in this study was defined as becoming more familiar with a passage. In future work it would be interesting t o study the development of a procedural skill from scratch. The effects of anxiety on attention in procedural m emory needs more research too. DeCaro, et al. (2011) explored the idea that anxiety causes people to focus too closely on a task, disrupting the procedural memory on which successful execution relies.
41 Experts were most successful when attention was dis tal and goal-oriented (Beilock, Carr, MacMahon, & Starkes, 2002; DeCaro, et al, 2011; Duk e, Cash, and Allen, 2011; Ford, Hodges, & Williams; McNevin, Shea, and Wulf, 2003). However, these do not directly measure anxiety's effects on people's focus while i n an anxious state, yet another opportunity for future work. Another open area of investigation is the compariso n of state-dependency across multiple procedural skills tested within the same f ramework. Perhaps state dependence varies in its effects across different kinds of ski lls. The current study explored the effects of acquiring procedural memory in an anxious state on recall of that memory in a similar state. Unfortunately, participants were not anxious enough to allow us to draw conclusions about the effects of this strategy on choking. However, because this knowledge could impr ove performance in sports, music, and typing, as well as many other common procedural tasks, the question is worth asking again in future work.
42 References Anderson, T. (1997, February). Contextual dependenc ies during perceptual motor skill activities: A re-evaluation of Wright and Shea (199 1). Dissertation Abstracts International, 57 Beilock, S. L., Carr, H. L., MacMahon, C., & Starke s J.L. (2002). When paying attention becomes counterproductive: impact of divided versus skill-focused attention on novice and experienced performance of sensorimotor skills. Journal of Experimental Psychology: Applied, 8 (1), 6-16. doi: 10.1037//1076-898X.8.1.6 Bond, C. F., & Omar, A. S.,(1990). Social Anxiety, State Dependence, and the Next-inLine Effect. Journal of Experimental and Social Psychology, 26, 185-198. Casasanto, D., & Dijkstra, K. (2010). Motor action and emotional memory. Cognition. 115 (1) 179-185, 179-185. doi: 10.1016/j.cognition.2009 .11.002 DeCaro, M. S., Albert, N.B., Thomas, R. D., & Beilo ck, S. L. (2011). Choking under pressure: multiple routes to skill failure. Journal of Experimental Psychology: General 140 (3) 390-406. doi:10.1037/a0023466 Duke, R.A., Cash, C.D., & Allen, S. (2011) Focus of attention affects performance of motor skills in music. Journal of Research in Music Education. Association of Music Education. Doi: 10.1177/0022429410396093 Duke, R.A., Simmons, A.L., & Cash, C.D. (2009). It s not how much; its how: Characteristics of practice behavior and retention of performance skills. Journal of Research in Music Education, 56(4), 310-321.
43 Ericsson, K. A., Krampe, R. T., & Tesch-Rmer (1993 ). The role of deliberate practice in acquisition of expert performance. Psychological Review, 100 (3), 363-406. Wright, D. L., & Shea, C. H. (1991), Contextual dep endencies in motor skills. Memory & Cognition, 19 (4), 361-370. Flanagan, J. R., & Johansson, R. S. (2003). Action plans used in action observation. Nature, 424 (14) 769-71. Ford. P., Hodges, N.J., & Williams M.J. (2005). Onl ine attentional-focus manipulations in a soccer-dribbling task: implications for the pr oceduralization of motor skills. Journal of Motor Behavior, 37( 5), 386-394. Gobel, E., Sanchez, D., & Reber, P. (2011) Integrat ion of temporal and ordinal information during serial interception sequence lea rning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37 (4) 994-1000. Hidalgo, V., Villada, C., Almela, M., Espn, L., G mez-Amor, J., & Salvador, A. (2012). Enhancing effects of acute psychosocial stress on p riming of non-declarative memory in healthy young adults. Stress, International Journal On The Biology Of Stress 15 (3), 329-338. doi: 10.3109/10253890.2011.624224 Immink, M. A., Wright, D.L., & Barnes, W.S. (2012) Temperature dependency in Motor Skill Learning. Journal of Motor Behavior, 44 (2) 103-113. doi: 10.1080/0022895.2012.64522 Lang, A. J., Craske, M. G., Brown, M., & Ghaneian, A. (2001). Fear-related statedependent memory. Cognition and Emotion, 51 (5), 695-703. doi: 10.108/0269930143000031
44 Lee, T. D., Magill, R. A. (1985). On the nature of movement representation in memory. British Journal of Psychology. 76 175-182. McCabe, R.E. (1999) Implicit and explicity memory f or threat words in highand lowanxiety sensitive participants. C ognitive Therapy and Research, 1 (23), 21-38. McDaniel, M.A., Anderson, C.D., Einstein, G.Ol, & O halloran C.M., (1989). Modulation of environmental reinstatement effects through enco ding strategies. American Journal of Psychology. 102 (4), 523-548. McNevin, H. H., Shea, C. H., & Wulf, G. (2003). In creasing the distance of an external focus of attention enhances learning. Psychological Research. Psychological Research. 67, 22-29. doi: 10.1007/s00426-002-0093-6 Mesagno, C., Harvey, J.T., & Janelle, M. C. (2012). Choking under pressure: The role of fear of negative evaluation. Psychology of Sport and Exercise. 13, 60-68. Doi: 10.1016/j.psychsport.2011.07.007 Miles, C. & Hardman, E. (1998). State-dependent mem ory produced by aerobic exercise. Ergonomics, 41 (1), 20-28. Opacic, T., Stevens, C., & Tillmann, B. (2009). Uns poken knowledge: Implicit learning of structured human dance movement. Journal Of Experimental Psychology: Learning, Memory, And Cognition, 35 (6), 1570-1577. doi:10.1037/a0017244
45 Park, J., Shea, C.H., & Wright, D.L. (2000). Reduce d-frequency concurrent and terminal feedback: A test of the guidance hypothesis. Journal of Motor Behavior, 32 (3), 287-296. Robinson, M. D., & Compton, R. J. (2006). The autom aticity of affective reactions: stimulus valence, arousal, and lateral spatial atte ntion. Social Cognition, 24 (4), 469-495. Shanks, D. R., & Johnstone, T. (1999). Evaluating t he relationship between explicit and implicit knowledge in a sequential reaction time ta sk. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25 (6), 1435-1451. Wright, D.L., Shae, C.H., & Whitacre Y. L. (1996) C ontextual dependencies during perceptual-motor skill acquisition: gone but not fo rgotten! Memory. 91-108. doi:10.1007/s00426-011-0388-6
47 Brief Survey 1. How interesting did you find the excerpt that yo u typed? Not at allVery much123452. Do you think that your typing skills have improv ed with college experience? Not at allVery much123453. Approximately how much typing do you do per week ? 0-2 pgs/week3-5-pgs/weekMore than 5 pgs/week4. Gender:5. Age:6. AOC:7. Year in college: