Extroversion, mental practice and imagery in the acquisition of the cartwheel

This work is aimed at (1) investigating the effects of physical practice (PP) added to mental practice (MP) with internal imagery (II) and external imagery (EI) on learning the cartwheel and (2) establishing correlations among scores of extroversion and task performances. Thirty female adolescents (mean age 12.93 ± 0.73 years) were assigned to one of two groups of PP followed by MP with either II (n=15) or EI (n=15). There were fi ve experimental phases: pre-test, acquisition, post-test, retention, and transfer. Movements were fi lmed in all phases, but the acquisition. A checklist was used to assess the movements and extroversion was evaluated by means of the Eysenck Personality Questionnaire. The analyses showed signifi cant intragroup improvements from the pre-test to the other tests but no intergroup differences were detected in any of the phases. Yet a positive and moderate correlation was detected between extroversion and performance in the tests for those who practiced with EI. These fi ndings reinforce the importance of MP as a complimentary tool to PP and show the existence of relationships between extroversion, cartwheel performance and imagery.


Introduction
Extroversion, mental practice and imagery in the acquisition of the cartwheel Mariana MOURA * Marina Mompean SOUZA * Jaqueline Freitas Oliveira NEIVA * Cássio Miranda MEIRA JUNIOR * (KI -in which the learner imagines movement sensations while picturing the skill) in learning an artistic gymnast sequence.Four experimental groups were created (EI, II, EI+KI and II+KI) and the sequence was relatively simple, composed of five continuous movements.Before practicing, each participant watched a video of the one of the four imagery conditions (from the group to which she/he was allocated).During acquisition, they pictured the steps of the sequence during six blocks of three sequences, with two minutes of rest in the interval of each block.Four weeks later, a retention test was performed, with the learners physically performing the sequence three times, with no instructions.To evaluate the performance, two referees graded the sequence with scores from 0 to 10 and the data analysis indicated that the EI+KI group received the higher grades in retention in comparison to the other groups.Th e diff erence in the fi ndings from the aforementioned studies is that the fi rst two [26][27] were performance studies while the last one 28 was a learning study which used tests of retention and transfer.In addition, there is a number of studies with mental practice and motor learning [29][30][31][32] that inspired the formulation of our hypotheses (last paragraph of the introduction section).Furthermore, our hypotheses incorporated the fact that external and internal imagery can be infl uenced by diff erent ways of imagining, as showed by questionnaires from studies which evaluated such ways of imagining [33][34] .
Individual differences can influence motor skill learning, and specifi cally the eff ectiveness of mental practice 18,35 .Th e performance in tasks has been associated to stable aspects of the individuals, named traces, which are permanent and consistent in several situations, predisposing individuals to act in a given way regardless of the circumstances.Th e focus of the present work is the individual diff erence of extroversion (E), a dimension of personality that is related to sociability and that has been an important characteristic to diff erentiate individuals in the motor domain [36][37][38][39][40][41][42] .Extroverts tend to be more sociable, sensation "hunters", expansive, enthusiastic, assertive and talkative; they sacrifi ce precision for velocity, and prefer to perform with other people short-term tasks that do not require attention to details.Rather, introverts tend to be shy, quiet, reserved and thoughtful; they sacrifi ce velocity for precision, prefer to work individually and can pay attention for longer periods.However, predisposition is not a certainty since extroverts and introverts do not always act in a given way; instead, they probably act in a given way [37][38][39][40][41]43 .
In the present study, E will be approached in the light of Eysenck's personality model [37][38] .A key concept in this model is arousal, which refl ects the state of being awoken, alertness or the corticopsychophysiological activation required for basic cortical process such as perception, attention, memory, learning and reasoning.Individuals diff er in the arousal baseline.For example, extroverts are chronically under-activated and are reactivated by high intensity stimuli.In turn, introverts are chronically over-activated and are reactivated by smooth and low-intensity stimuli 44 .
Th ese diff erences that defi ne distinct personality traits between persons have been associated to performance in motor tasks.Th e eff ects of E in the motor domain have been linked to issues of sequential movement planning and execution [45][46][47][48][49][50][51] , target accuracy 52 , speed-accuracy trade-off [53][54][55] and practice schedule in motor learning 55 .Th e present study contributes to the knowledge on the relation between E and skill acquisition by investigating MP with distinct perspectives of imagery.Our main purpose was to study the effects of PP associated with MP with II or EI in learning the cartwheel.Given that extroverts and introverts have diff erent levels of cortical activation, we sought to determine relationships between E scores and task performance.According to Hardy and Callow 28 , we tested the hypothesis that PP combined with MP through EI would result in better retention and transfer cartwheel scores.Furthermore, provided the particular characteristics of introverts (introspective, self-centered and more focused on the task details) and extroverts (externally-centered, and more focused on the general aspects of the task), we expected a strong negative correlation between E and retention/transfer performance when MP was internally centered (II), but a high positive correlation between the same scores when MP was externally centered (EI) 37,41 .

Sample
Th irty female adolescents (12.93 ± 0.73 years) with no experience in the task volunteered to participate in the study.The study rigorously followed the Declaration of Helsinki on human research ethics -version VII (2008), published by the World Health Organization.Th e participation was authorized by parents/tutors, who signed the informed consent form.

Task, apparatus and procedures
Th e Eysenck Personality Questionnaire (EPQ)validated for Brazilian population 56 -is composed of 88 questions and was used to quantify E levels.Th e variation amplitude in the score of this trait is from 0 to 18.
Th e cartwheel was chosen as the experimental task, which consists of six submovements described as follows 6 : 1) upright position with arms high in the air and straight, step forward; 2) push off of the front leg; 3) place hands on the ground; 4) kick legs up and over the torso and head as the body goes inverted; 5) legs stay apart during the rotation; set one foot on the ground, followed by the other foot, landing in a lunge; 6) fi nish the movement in the upright position with arms/hands straight and high, and legs apart.
Four one-minute videos were recorded with the movements of an experienced gymnast performing the task.Two videos represented the internal perspective, recorded with the camera fi xed on the forehead with a special helmet in the model's head.Th e other two videos represented the external perspective, with the recording being performed with a camera positioned 2 meters away, laterally to the model.One video of each perspective was recorded with the cartwheel from left to right and the other from right to left.Th e videos were used to familiarize participants with the task at the beginning of the experiment and to instruct them about the key points of the task during the acquisition phase.Participants were allowed to watch the videos as many times as they wished.
During the experiment, the recordings were performed with a SONY HDR-PJ340 camera placed on a tripod 2 meters away from the movement execution.A 4 meters long straight line was fi xed on the fl oor with adhesive tape to help participants to orientate themselves.
A score was given to each movement according to the movement pattern evaluation checklist from Velloso 57 , resulting in four values in the pre-test, four values in the posttest, four values in retention and four values in transfer.Th e scores varied between 3 and 10 and were attributed by two specialists in artistic gymnastics (intra and interobserver reliability are presented in the data analysis section).Th e checklist was as follows: Write "x" in one of the three items of evaluation for each movement phase.Participants were instructed to sit in an armchair, with eyes closed, and to perform the MP of the cartwheel with either II or EI (depending on the group to which the participant was allocated).In addition, the participants received the instruction that the MP must focus clearly on the cartwheel, which must be mentally practiced as vividly as possible, with real sequential and temporal aspects of the movement so that to obtain a successful mental trial.

Experimental design
Participants were randomly assigned to PP + MP groups: Th e II group (GI) started with PP and then performed MP with II (n=15), while the EI group (GE) started with PP and then performed MP

Results
Cronbach's  indicated very reasonable reliability of the instrument.The intraclass correlations were =0.967 (p<0.0001) for the fi rst observer and =0.947 (p<0.0001) for the second one.Th e interobserver correlation was =0.902 (p<0.0001).
Performance in each phase are represented by the scores (mean ± standard deviation) of each group showed in TABLE 1. of acquisition 3.Each participant performed the posttest following the same procedures of the pretest.In the third day, participants performed the retention and transfer tests.In the retention test, participants performed physically four trials of the complete cartwheel with their preferred side, with no feedback.After ten minutes, transfer test was run with participants being submitted to the same retention procedure, but performing with the nonpreferred side.All trials were video recorded.

Data analysis
Reliability of the checklist 57 was performed via intra and interobserver correlations (Cronbach's 58 ).Ten movements recorded were randomly selected, three from pre-test, three from posttest, two from retention test, and two from transfer test.Th e observers analyzed each set of trials twice, with an interval of 20 days between them.
The scores for each recorded movement were compared within and between groups through descriptive (mean and standard deviation) and inferential statistics (pre/posttest and retention/ transfer tests).For each step of the inferential analysis, we used a two-way [2 (groups) x 2 (tests)] analysis of variance (ANOVA). 2 values were reported to indicate effect sizes for the significant results.
Relationships between extroversion and type of mental imagery were established with the use of Pearson correlation's (r) for each test per group.To achieve this goal, we used the scores of each personality trait and the task performance.For all analyses, alpha was set at 5%.Statistical procedures were performed on SPSS 17.0.
with EI (n=15).Th e experiment consisted of the following phases: pre-test, acquisition 1, acquisition 2, acquisition 3, posttest, retention and transfer.We opted for using PP before MP based on the theoretical idea that one must have PP before MP 23 .
In the pre-test, participants learned about the task by visual instruction and by watching a video depicting a model performing the skill from an external perspective with the preferred side of the learner.Along with the video, an audio highlighted relevant aspects of the task.Th e one-minute video was reproduced five times.Next, participants physically performed four complete trials of the cartwheel using their preferred side.
In acquisition 1, the identical visual instruction used in the pre-test was provided for both groups.Th en, each participant performed two blocks of 20 physical repetitions, with a fi ve-minute resting interval between blocks, during which she watched the same instructional video.Acquisition 2 was administered on the same day as acquisition 1, with participants of GE watching the pre-test video; for GI's participants, the video was shown from an internal perspective, with the model performing with a camera fi xated on the forehead so that to give the impression that the skill is being performed in the fi rst person.GI's participants were informed that the MP must be performed in the fi rst person (II), whereas GE's participants informed that the MP must be performed in the third person (EI).After watching the video, the GE group mentally practiced with EI and the GI group with II.Both groups performed 40 trials (20 trials, fi ve-minute rest, 20 trials).
Acquisition 3 took place in the second day using the same procedure as acquisition 2. Th e posttest was performed fi ve minutes after the end ANOVA did not identify significant differences between groups.Nevertheless, a significant main effect for "Block", from pre to posttest [F(1,28)=105.99;p=0.002;  2 =0.79], was detected.This performance change can be interpreted as an indicative of learning 1 .No significant effect was found from posttest to retention (persistence), whereas there was a significant difference between posttest and transfer (adaptability), regardless of group [F(1,28)=14.95;p=0.001;  2 =0.35].
In GE group, the retention correlation between extroversion and performance scores was significant, negative and moderate (r=-0.61;p=0.016).In the transfer test, the correlation also reached significance, but of low magnitude (r=-0.46,p=0.023).No significant relationships were found for the GI group.

Discussion
Th e main purpose of the present study was to investigate the imagery perspective, internal or external, during MP combined with PP in the cartwheel learning.We expected that PP with MP through EI would lead to better scores in both retention and transfer.However, we did not detect significant differences between groups in any of the experimental phases.Th e hypothesis, set according to a study with adult women practicing artistic gymnstics 28 , was refuted.It seems that the development level of the participants, as well as the task complexity, were responsible for distinct results in both studies.Despite the fact that both tasks can be categorized as serial skills (chain of discrete movements) 1 , the cartwheel bears fewer components when compared to the fl oor event.Th e developmental diff erence of the participants (adolescents x adults) might be another factor for distinct eff ects in terms of imagery perspective.It is worth noting that in Hardy and Callow's 28 study, the EI was eff ective only when the learner imagined the sensations of the movement; this was not the case in the present study because our participants were not instructed to employ kinesthetic imagery in any of the experimental groups.Future research should address this issue.
Even though we could not detect significant diff erences between groups, there were intragroup diff erences for both groups.Both internal and external imagery groups showed remarkable increase in performance from pre to posttest.Th is improvement has been interpreted as one of the criteria for inferring motor learning.Another criterion for learning occurrence is persistence, refl ected in the lack of diff erences between posttest and the retention test (performed 24h later) 1 .However, the change in the task employed in the transfer test (performed with the non-preferred side) caused signifi cant performance decrement.Th erefore, MP + PP enhanced cartwheel's performance, regardless of imagery perspective (fi rst or third person).
Th e benefi cial cognitive eff ect of mental rehearsal is that the learner draws attention to relevant aspects of the target-skill so that information processing mechanisms are strengthened.These cognitive mechanisms might facilitate the means to reach the goals of the target-skill, as well as to consolidate strategies for error detection and correction in more advanced stages 20 .Also, MP yields the activation of muscles that would be responsible for the real execution of the movement.In fact, some studies reported brain activation in areas which are responsible for the real movement during MP 15,59 .Th ese neural and cortical explanations are grounded on the idea that mental images and their correspondent mental representations share the same properties and, consequently, play the same role of PP in generating motor representations [16][17] .It is recommended thus to use PP combined with MP, yet to optimize PP time in a practice session.
Th e order between PP and MP during practice has generated debate.Th e argument for "MP before PP" highlights that MP eff ects might be hindered by previous PP 60 .Mental rehearsal would overload the learner's capacity to extract relevant information from the task and, as a result, the learner would be engaged in positive higher cognitive eff ort (or demand) 61 .However, this hypothesis is in contrast to previous literature [10][11][62][63] . In the resent study, PP was performed prior to MP, in line with the assumption that one should fi rst experience the real execution of the movement as a prerequisite for MP 23,64 .In summary, the implementation of PP fi rst seems to optimize the eff ects of MP, probably because PP (1) entails relationships between error detection/ correction and intrinsic feedback 1-2 , (2) does not require conscious control of the action 15,[61][62][63][64] , and (3) provides relevant aff erent organization to integrate perception and action [65][66][67][68][69] .As these aspects are absent from MP, this type of practice is less eff ective than PP.Nevertheless, MP is more eff ective than no practice whatsoever and it seems to enhance learning when implemented after PP.
The secondary goal of the present study was to describe the significance and the magnitude of correlation between E scores and groups' performance.A discrepancy was detected between correlations' values/directions obtained and what was expected.Retention and transfer correlations were signifi cant, negative and moderate between E scores and performance only for those who mentally practiced in the third person.The correlations' direction for this group was opposite to the direction stated in our hypothesis: lower values of E (introversion) were associated with higher performance, while higher values of E (extroversion) were associated to poorer performance.According to the particular characteristics of extroverts and introverts, higher E scores would be associated to better performance with EI.Rather, our fi ndings give evidence as to the opposite.It is theoretically arguable that our introverted participants, because they are over-activated (higher levels of basal arousal), were self-reactivated to optimal points of performance with EI, an imagery perspective that might have been perceived by them as a lower intensity stimulus 44 .It also appears plausible to believe that EI must have served as a useful stimulation eff ect for introverts to engage more actively in the tests.
In conclusion, the present findings gave evidence to the eff ect that there is a negative correlation between E scores and the cartwheel performance when, next to PP, MP is performed through external imagery.

TABLE 1 -
Mean ± standard deviation of the grades obtained for each group (GI: internal imagery group; GE: external imagery group) in each experimental phase.