Perceptual and Motor Skills, 2010, 110, 3, 888-896. © Perceptual and Motor Skills 2010

DOI 10.2466/PMS.110.3.888-896 ISSN 0031-5125

MENTAL PRACTICE AND THE RETENTION OF MOTOR LEARNING: A PILOT STUDY1

MICHAEL SPITTLE

School of Exercise and Nutrition Sciences

PETER KREMER

School of PsychologyDeakin University, Victoria, Australia

Summary.—This study compared the effects of mental practice, physical prac-tice, and no practice on the performance and retention of a novel throwing task. Research supports the effectiveness of mental practice on performance; however, retention of learning has not been adequately investigated. Participants were 152 students ages 18 to 44 years (M = 20.5, SD = 2.9), who completed a pretest, posttest, and five-week delayed retention test of dart throwing with the nonpreferred hand. In the practice phase, participants completed 50 mental practice or physical practice trials of the darts task or 50 trials of a catching task. Results indicated that overall scores increased from pretest to posttest and retention test, and decreased from posttest to retention test, but that these effects did not differ for type of practice. The findings suggest equal learning and retention of learning for novel throwing tasks for control, mental practice, and physical practice conditions; however, further re-search that considers issues raised is recommended.

Mental practice refers to the mental repetition of a task, without ob-servable movement, with the specific intent of learning (Corbin, 1972), that is, practicing the skill “in your head.” It generally involves using im-agery or some other cognitive process to practice a skill repetitively (Mor-ris, Spittle, & Watt, 2005). There is a large body of research that supports the effectiveness of mental practice, with research reviews consistently re-porting that mental practice produces larger changes in learning than no practice, but smaller changes than physical practice (Weinberg, 1982; Feltz & Landers, 1983; Feltz, Landers, & Becker, 1988; Hinshaw, 1991; Grouios, 1992; Driskell, Copper, & Moran, 1994). The traditional mental practice re-search design involves a simple pretest to posttest design, comparing the effects of mental practice to no practice and physical practice on the per-formance of a skill (Morris, Spittle, & Perry, 2004). The most common de-sign involves a pretest of performance, followed by mental practice of 10 to 20 minutes’ duration, then a posttest of performance, usually measur-ing performance on an unfamiliar artificial analogue task. This study at-tempts to recreate much of this typical design, but to further investigate retention.

Learning is a change in the capability to perform a skill that must be inferred from a relatively permanent change in performance (Magill,

1Address correspondence to Dr. Peter Kremer, School of Psychology, Deakin University, Wa-terfront Campus, Geelong, Victoria, Australia 3217 or e-mail (peter.kremer@deakin.edu.au).

MENTAL PRACTICE AND RETENTION 889

2007), not just a short-term improvement in performance. Measuring per-formance changes in one 30-minute session, as is common in pretest to posttest designs, is not adequate to assess a relatively permanent change in performance. This permanence or persistence of learning means that there should be an improved level of performance today, tomorrow, next week, next month, and so on (Magill, 2007). This is often assessed through the use of retention tests, where the participant is retested some time after initial learning and after a period without practice of the skill, but long-term retention has rarely been explored with mental practice (Murphy & Martin, 2002; Morris, et al., 2005).

A meta-analytic review of the general skill-learning literature report-ed that there is a substantial loss of skill learning when there are extend-ed periods of nonpractice and the amount of skill decay increased as the length of the retention interval increased (Arthur, Bennett, Stanush, & Mc-Nelly, 1998). Whether this skill loss applies to skills learned using mental practice rather than physical practice is relatively unexplored, as is wheth-er skills learned using mental practice are more or less resistant to decay than skills learned through physical practice. Although a few studies have reported retention scores to examine mental practice, the retention inter-vals utilised have generally been short, such as 10 minutes (e.g., Overdorf, Schweighardt, Page, & McGrath, 2004). For example, in their meta-analy-sis Driskell, et al. (1994) reported only eight mental practice studies with retention intervals of more than 1 day, with seven of these studies using retention intervals of 10 days or less (Clark, 1960; Gondola, 1966; Cronk, 1967; Hackler, 1971; Noel, 1980; Puretz, 1987; Hird, Landers, Thomas, & Horan, 1991). The exception to this was a study by Kovar (1967) in which participants learned an underhand free throw task and were assessed on accuracy of performance immediately after practice, and after 28 and 35 days. Kovar found no significant differences between mental and physi-cal practice. One study not reported by Driskell, et al. (1994), conducted by Gabriele, Hall, and Lee (1989), used a 2-week retention test and reported that random imagery practice was as beneficial to retention as random physical practice.

Driskell, et al. utilised regression to estimate the effect of retention interval on the effects of mental practice. The strongest effect of mental practice was for immediate retention, with an expected effect size using Fisher’s Z of .33, which fell to .26 at 7 days, .19 at 14 days, and .10 at 21 days. The retention interval had less estimated effect on performance af-ter physical practice, with immediate retention of .52, which decreased to .45 at 7 days, .38 at 14 days, and .31 at 21 days. Driskell, et al. (1994) con-cluded that the retention interval did moderate the effectiveness of men-tal practice, with longer retention intervals producing weaker effects on

M. SPITTLE & P. KREMER890

performance. It should be emphasised that these predictions are estimates based on only seven studies; consequently, the effect of the length of the retention interval on learning through mental practice remains unclear. The retention intervals investigated in the general practice literature have often exceeded those used in the mental practice studies (Arthur, et al., 1998), so studies exploring longer retention intervals with mental practice are clearly needed.

This study recreated the typical mental practice design, but also incor-porated a retention test to examine the long-term effects of mental prac-tice on the learning of a motor skill. It was expected that physical practice would lead to greater improvement in performance from pretest to post-test than mental practice and that improvements for both practice groups would be greater than for a control group who completed an unrelated catching task. It was also expected that the effects for the physical practice condition would be better retained at the retention test than for the men-tal practice condition, and again the effects for both practice conditions would be better retained than for the control condition.

MethodParticipants

Participants were 152 undergraduate students, 74 men and 78 women, recruited from university courses in human movement, physi-recruited from university courses in human movement, physi-cal education, and psychology who were between 18 and 44 years of age, with a mean age of 20.5 yr. (SD = 2.9). The sample was a convenience sam-ple of volunteers. A short questionnaire was used to gather demographic information and details about participants. Participants rated their darts ability as elite/advanced/competitive (n = 28), noncompetitive (n = 39), be-ginner (n = 45), or novice/never played darts (n = 37); responses for three participants were missing. Most participants reported their previous ex-perience of mental practice was limited to learning about it in class (n = 81), while a small number had formal training in imagery (n = 34). Equipment

The dart-throwing task involved throwing darts at a dartboard with concentric circles from a distance of 2.44 m. The diameters of the concen-tric circles ranged from 1.5 cm to 46 cm. The dartboard was fixed with the centre of the target set at a height of 160 cm. Distance and location of the dartboard were assessed by pilot work with three individuals to manipu-late the difficulty of the task. The aim in presetting the difficulty was for naive performers to achieve a score of approximately 50% of maximum at pretest, thus creating a sufficiently difficult task so there would be ad-equate opportunity for improvement due to practice at posttest.

MENTAL PRACTICE AND RETENTION 891

ProcedureEthical approval for the study was provided by the University Hu-

man Research Ethics Committee. Testing was completed in two sessions, each approximately 50 min. in duration. During the first session, partici-pants were assigned to one of three practice conditions. Participants com-pleted a pretest consisting of 50 trials of dart throwing divided into 10 blocks of five throws each with the nonpreferred hand. Participants then performed 50 practice trials according to their assigned condition before completing the posttest following the same procedure. The second ses-sion was the retention test (same test procedure), which occurred 5 wk. after the initial session. Scores achieved for each trial on the three perfor-mance tests as well as the practice trials were recorded. For darts hitting the centre circle, participants scored 10 points, the next circle out scored nine points, and so on. Darts hitting the region beyond the outermost cir-cle were scored as 0. This gave an accumulated score with a range of 0 to 50 for each block of trials. The skill was self-paced in that participants could throw whenever they were ready.

Procedures for the practice phase for the three conditions were as fol-lows. In the Physical Practice Condition, participants performed 50 physi-cal practice throws in 10 blocks of five throws. In the Mental Practice Con-dition, participants performed 50 mental practice trials in 10 blocks of five throws. Participants were instructed to “Imagine throwing the dart at the centre of the dartboard, feel the movement and imagine successful-ly hitting the middle of the target.” Participants recorded their imagined scores for each trial. In the No Practice condition, participants performed no practice of the dart-throwing task. Instead they performed a one-hand-ed catching task of a standard tennis ball for 50 trials in 10 blocks of five catches with a partner across a distance of 2.5 m. The ball was thrown un-derhand to the participant. Participants recorded the number of success-ful catches.Design, Measures, and Analysis

An experimental research design was used which incorporated a pre-test, posttest, and retention test comparison of the practice conditions. The averaged total score achieved over the 10 blocks of trials on the dart-throwing task was the performance measure. Mean pretest, posttest, and retention test performances were computed. Three separate (subject vari-ables: Darts Ability, Previous Mental Practice, Sex) independent-samples t tests were used to test for any pre-existing differences that could con-found interpretation of effects. For two of these analyses, the number of categories was reduced to two to avoid problems with low cell numbers. Where significant effects were found, post hoc tests were conducted to as-sess whether the variables were equally distributed across the three prac-

M. SPITTLE & P. KREMER892

tice conditions using χ2 tests for independence. Multivariate repeated-measures analysis of variance (MANOVA) was used to examine effects for type of practice (Practice) across the three tests. The multivariate ap-proach overcomes problems associated with the assumption of sphericity (Stevens, 1986). Two transformed variables were generated using orthogo-nal contrasts on scores across the three tests. The first variable represented the change from pretest to the combination of posttest and retention test, and the second variable represented the change in scores from posttest to retention test. Two orthogonal contrasts were used to test specific effects for type of practice: (1) the two practice conditions against the No Practice condition, and (2) the Physical Practice condition against the Mental Prac-tice condition. Multivariate tests were evaluated using Wilks Lambda (λ) and follow-up univariate tests using the t statistic. All analyses were per-formed using SPSS Version 17.0.

ResultsMean performance scores for pretest, posttest, and retention tests

for the three practice conditions are presented in Table 1. Results of the separate t tests showed that pretest scores did not differ for previous mental practice or training but did for darts ability (t147 = 4.8, p < .001) and Sex (t150 = 6.5, p < .001). Follow-up analyses revealed that neither darts ability [χ2(n = 149) = 4.5, p > .05] or Sex [χ2(n = 152) = 3.1, p > .05] was associated with practice condition, and that darts ability and sex were equally distributed across groups.

TABLE 1Mean Pretest, Posttest, and Retention Test Scores

(Points) With Standard Errors by Practice Condition

Condition n Pretest Posttest Retention TestM SE M SE M SE

Physical Practice 51 24.10 0.65 26.45 0.58 25.24 0.60Mental Practice 55 24.73 0.67 25.92 0.72 25.68 0.66No Practice 46 25.41 0.75 26.67 0.76 25.80 0.71

MANOVA revealed significant effects for Test (F2,298 = 14.0, p < .001; ηp

2 = 0.09), but not for Practice or the Test × Practice interaction (Fs < 2.0, p > 05). Multivariate results for the two transformed measures revealed significant effects for Test (Wilks λ = 0.8, p < .001; ηp

2 = 0.07) but not for the Test × Practice interaction (Wilks λ = 1.0, p > .05; ηp

2 = 0.01). Univariate results for the first measure indicated that the combination of scores at posttest and retention test was higher than pretest scores (t = 4.8, p < .001; Cohen’s d = 0.25) but this effect was not different for the combination of the two practice conditions compared with the No Practice condition, or the Physical Practice condition compared with the Mental Practice condition (ts < 2.0, p > .05). Results for the second measure indicated similar effects.

MENTAL PRACTICE AND RETENTION 893

Retention test scores were lower than posttest scores (t = −2.4, p < .05; Cohen’s d = 0.15) but this effect was not different for the combination of the two practice conditions compared with the No Practice condition, or the Physical Practice condition compared with the Mental Practice condition (ts < 2.0, ps > .05).

DiscussionIn this study, the long-term learning of a simple motor skill was in-

vestigated under conditions of physical practice, mental practice, and no practice. Overall, performance improved from pretest to posttest and re-tention test, and performance declined from posttest to retention test. The magnitude of the improvement and then the decline, however, did not dif-fer by type of practice.

The findings failed to show superior improvements for either the Physical Practice or Mental Practice conditions over a No Practice con-trol condition. This finding is inconsistent with previous studies that have reported that physical practice is superior to mental practice and mental practice is superior to no practice for motor skill learning (Feltz & Land-ers, 1983; Feltz, et al., 1988; Driskell, et al., 1994). In the current study, the No Practice condition realised some improvement from pretest to post-test. It must, however, be noted that in previous research no practice has often had an effect on performance. For instance, Feltz, et al. (1988) report-ed larger effect sizes for physical practice (0.72) and mental practice (0.47); however, there was still an effect for no practice (0.22). It is also possible that the No Practice condition benefited from the 50 pretest trials, which may have constituted some physical practice of the skill. Designs where participants are matched on key measures and then complete the same practice conditions, but without any baseline pretest, could overcome this problem. More complex designs that incorporate the structure of the Solo-mon design (Kerlinger & Lee, 2000) could be used. An alternative explana-tion is that the catching task may have engaged the participants in percep-tual and motor activities that kept them in a perceptual-motor set, which was reflected in the posttest scores, but was not evident at retention. This finding further supports the need to test the longer term effects of differ-ent practice conditions.

The major purpose of the study was to evaluate the effect of mental practice on retention. The findings showed that type of practice did not influence results on the posttest-to-retention-test measure. These findings are also in contrast to the limited previous literature, which has suggested that physical practice is likely to lead to more effective retention of learn-ing than mental practice (Driskell, et al., 1994). Arthur, et al. (1998) report-ed substantial skill loss for nonpractice after initial practice in the general practice literature, so decreases after physical practice would be expected.

M. SPITTLE & P. KREMER894

Findings showed that the decrease for the mental practice condition was equal to that observed for the physical practice condition.

The darts task required participants to throw with the nonpreferred hand, a task that was likely to be novel to them. Cross-transfer (bilateral transfer) from one limb, however, could occur in the current task, given that some participants will have had experience in throwing a dart with their preferred hand. Despite this, there was no difference in mean scores between participants who reported that they had previous preferred hand dart-throwing experience and those having little or no previous experience at pretest. This suggests that there was actually little cross-hand transfer or that participants were not accurate in reporting their experience.

It could be argued that the lack of a strong learning effect was due to low amounts of practice, that is, only 50 practice trials. However, pre-vious reviews (e.g., Feltz & Landers, 1983) and separate studies (e.g., Et-nier & Landers, 1996) have highlighted uncertainty about the relationship between the amount of mental practice required for learning effects and have often suggested that small amounts of mental practice may produce larger effect sizes than performing more mental practice (Feltz & Land-ers, 1983; Hinshaw, 1991; Etnier & Landers, 1996). A recent study by Kre-mer, Spittle, McNeil, and Shinners (2009) reported that groups who com-pleted either 25, 50, or 100 mental practice trials of a dart-throwing task with the nonpreferred hand all improved similarly from pretest to post-test. Perhaps small amounts of practice are sufficient to bring about short-term changes in performance, but more practice is needed to bring about changes in long-term retention of learning. Further research on retention that includes greater amounts of practice is recommended.

This study found similar effects for mental practice, physical practice, and no practice from pretest to the combination of posttest and retention tests, as well as no differences in the decline of learning at retention test-ing. Further research that incorporates a number of design modifications is needed to further investigate the relationship between type of practice and retention of learning.

REFERENCES

Arthur, W., Bennett, W., Stanush, P. L., & McNelly, T. L. (1998) Factors that influ-ence skill decay and retention: a quantitative review and analysis. Human Perfor-mance, 11, 57-101.

Clark, L. V. (1960) Effect of mental practice on the development of a certain motor skill. Research Quarterly, 31, 560-569.

Corbin, C. B. (1972) Mental practice. In W. P. Morgan (Ed.), Ergogenic aids and muscular performance. New York: Academic Press. Pp. 94-118.

Cronk, J. M. (1967) The effect of physical practice, mental practice, and physical-men-tal practice on the development of arm strength. Unpublished doctoral disserta-tion, Florida State Univer., Tallahassee.

MENTAL PRACTICE AND RETENTION 895

Driskell, J. E., Copper, C., & Moran, A. (1994) Does mental practice enhance perfor-mance? Journal of Applied Psychology, 79, 481-492.

Etnier, J. L., & Landers, D. M. (1996) The influence of procedural variables on the efficacy of mental practice. The Sport Psychologist, 10, 48-57.

Feltz, D. L., & Landers, D. M. (1983) The effect of mental practice on motor skill learning and performance: a meta-analysis. Journal of Sport Psychology, 2, 211-220.

Feltz, D. L., Landers, D. M., & Becker, B. J. (1988) A revised meta-analysis of the mental practice literature on motor skill learning. In D. Druckman & J. Swets (Eds.), Enhancing human performance: issues, theories, and techniques. Washington, DC: National Academy Press. Pp. 1-65.

Gabriele, T. E., Hall, C. R., & Lee, T. D. (1989) Cognition in motor learning: imagery effects on contextual interference. Human Movement Science, 8, 227-245.

Gondola, J. C. (1966) A comparison of the effectiveness of programs of physical prac-tice and mental practice on the performance of a selected test of balance. Unpub-lished doctoral dissertation, Purdue Univer., West Lafayette, IN.

Grouios, G. (1992) Mental practice: a review. Journal of Sport Behavior, 15, 42-59.Hackler, C. M. (1971) An experimental study of the relative effectiveness of physical

and mental practice on the learning and retention of a selected motor task. Un-published doctoral dissertation, Univer. of Maryland, College Park.

Hinshaw, K. E. (1991) The effects of mental practice on motor skill performance: criti-cal evaluation and meta-analysis. Imagination, Cognition and Personality, 11, 3-35.

Hird, J. S., Landers, D. M., Thomas, J. R., & Horan, J. J. (1991) Physical practice is superior to mental practice in enhancing cognitive and motor task performance. Journal of Sport and Exercise Psychology, 13, 281-293.

Kerlinger, F. N., & Lee, H. N. (2000) Foundations of behavioral research. (4th ed.) Fort Worth, TX: Harcourt College.

Kovar, S. V. (1967) The relative effects of physical, mental and combined mental prac-tice in the acquisition of a motor skill. Unpublished master’s thesis, Univer. of Illinois, Urbana.

Kremer, P., Spittle, M., McNeil, D. G., & Shinners, C. (2009) Amount of mental prac-tice and performance of a simple motor task. Perceptual and Motor Skills, 109, 347-356.

Magill, R. A. (2007) Motor control and learning: concepts and applications. (8th ed.) Dubuque, IA: McGraw-Hill.

Morris, T., Spittle, M., & Perry, C. (2004) Mental imagery in sport. In T. Morris & J. Summers (Eds.), Sport psychology: theory, applications, and issues. (2nd ed.) Brisbane, Australia: Wiley. Pp. 344-383.

Morris, T., Spittle, M., & Watt, A. P. (2005) Imagery in sport. Champaign, IL: Human Kinetics.

Murphy, S. M., & Martin, K. A. (2002) The use of imagery in sport. In T. S. Horn (Ed.), Advances in sport psychology. (2nd ed.) Champaign, IL: Human Kinetics. Pp. 405-439.

Noel, R. C. (1980) The effects of visuo-motor behavior rehearsal on tennis perfor-mance. International Journal of Sport Psychology, 2, 221-226.

Overdorf, V., Schweighardt, R., Page, S. J., & McGrath, R. E. (2004) Mental and phys-ical practice schedules in acquisition and retention of novel timing skills. Percep-tual and Motor Skills, 99, 51-62.

M. SPITTLE & P. KREMER896

Puretz, S. L. (1987) Movement retrieval from long-term memory: physical versus mental rehearsal. Perceptual and Motor Skills, 65, 603-612.

Stevens, J. (1986) Applied multivariate statistics for the social sciences. Hillsdale, NJ: Erl-baum.

Weinberg, R. S. (1982) The relationship between mental strategies and motor perfor-mance: a review and critique. Quest, 32, 195-213.

Accepted May 4, 2010.