Generalized motor program

A generalized motor program is thought to develop over practice and provides the basis for generating movement sequences within a class of movements that share the same invariant features, such as sequence order, relative timing, and relative force. Specific movements are produced by the premovement specification of movement parameters like absolute timing, absolute force, and effectors.

For example, when a movement situation requires a learned sequence to be produced either faster or slower than typically practiced, the invariant features remain unchanged, but the absolute timing parameter is changed to accommodate the rate at which the movement is produced.

Thus, the relative times used to produce the elements in the sequence remain unchanged, but the absolute time is rescaled to meet the specific demands. Changing the absolute timing parameter results in slower movements that could be considered stretched out in time copies of faster movements. Likewise, a lower specification of the force parameter would result in a movement sequence generating reduced forces, which could be thought of as a compressed with respect to force copy of a more forceful movement.

Each of us can write our signature under a variety of conditions. According to the GMP perspective, we do this by specifying the movement parameters needed to meet the requirements at hand while maintaining the invariant features of the GMP. For example, if I were asked to write my name Charles Shea very quickly on a sheet of paper, the signature maintains the relative timing characteristics invariant to the GMP, but the specification of absolute timing would be reduced.

This will result in me taking the same proportion of the total time to write the Ch , for example, in my first name when writing quickly as when writing under normal time constraints, but the absolute time used would be reduced. Similarly, if I were asked to write my name in a small box on a sheet of paper or much larger on a white board in the classroom, the invariant features would not change, but the actual timing, actual forces, and even the specific effectors used to produce the movement would change.

In the smaller situation, one would use primarily finger movements to produce a signature, while in the larger situation one may use shoulder and arm movements with minimal or no movements of the fingers.

The notion of a GMP has a great deal of intuitive appeal. It seems efficient for each motor program in our movement repertoire to be able to generate a class of movement sequences. This reduces not only the potential storage problems that would result if different programs were needed each time the movement requirements changed but also would reduce potential retrieval problems that would be associated with selecting from among a group of similar motor programs.

Participants were instructed to throw the ball like in the basketball free throws. During pretest and posttests, the first five shots at the distance 4. A 7 Hz low pass filter with 2nd order low-pass was used and an optimal cut-off frequency based on the residual analysis was applied Winter, The system had been calibrated statically and dynamically before recording started and was re-calibrated during testing if environmental factors changed e.

A total of nine reflective markers were placed on the skin on the dominant side of the participant: distal end of the fifth metatarsal of the toe, lateral malleolus of the ankle, lateral condyle of the femur and on the greater trochanter of the femur, distal end of the middle finger just below the nail, on the hand just below the middle finger, ulnar styloid of the wrist, lateral epicondyle of the elbow and on the acromion process of the shoulder.

Participants wore sleeveless shirts or no shirts. Since the greatest angular range of motion during the propulsion phase of the shot can be noticed in elbow Button et al. Recordings, as well as practice sessions, took place on the university basketball court. The relative time as a percentage of overall movement time to reach five kinematic landmarks in the elbow joint was analyzed Schneider and Schmidt, ; Breslin et al. Figure 1. Landmarks A-E peak velocity-A, peak flexion-B, peak acceleration-C, peak velocity-D and negative peak acceleration -E on an exemplary graph representing one of the participants: A — angle, B — angular velocity, C — angular acceleration.

The lines were smoothed using the moving average based on 20 following results of the full data set, including the one representing particular time point.

Bonferroni adjustments for multiple comparisons were applied. We used a previously applied method to detect especial skills Keetch et al. Average percentage accuracy scores based on a 2-point scoring system were calculated for all the trials across each of the seven distances. We used these scores to compute regression lines for all of the shooting distances but the free-throw distance 4.

Based on regression equations, we estimated predicted shot efficiency at the distance of 4. We calculated the relative time as a percentage of overall movement time to reach five kinematic landmarks. In the VP group, all kinematic landmarks occurred relatively later in posttest than in pretest see Table 1 for mean values.

We computed a linear mixed model as described above. We did not find any significant main i. Table 2. Results of the linear mixed model for Experiment 1: F -values, p -values, and ES — effect sizes. Main effects group: CP vs. VP; test: pretest vs. Effect sizes for the interaction comparisons landmark B were small VP group in pretest vs. We used the same model as in the previous analysis for analyzing movement time MT Table 3. The only significant effect was the group effect.

Table 3. Results of the linear mixed model with the movement time MT as a dependent variable. F -value, p -value and ES — effect size. We found that the MT in VP lengthened in posttest as compared to the pretest. The reverse effect was observed in CP group — MT in posttest was shorter than in pretest.

Table 4. ES — effect size. We had two objectives in this study: to determine whether CP conditions develops different GMP as opposed to the variable practice conditions. Secondly, we expected an especial skill to emerge in a limited number of trials in CP conditions.

In regards to the latter objective, we were unsuccessful at developing an especial skill in CP. Although our participants performed more shots during practice than participants in Breslin et al.

We may speculate that because we used a different method of scoring, i. Perhaps, the 4-point scoring system is more sensitive and therefore, we would have found an especial skill if we had used it. We found no difference between CP and VP in posttests and the only significant interaction effect was found in landmark B. Given, that the only significant differences between landmarks were found between pre- and posttests results we could assume that CP and VP groups developed the same GMP governing shots at the distance 4.

Unfortunately, the significant interaction effect found in landmark B blurred our results. As Schneider and Schmidt pointed out, relative timing is essentially invariant, should a unit of action be governed by a single GMP. Therefore, we cannot claim whether VP develops a different GMP than the CP but at the same time, we should be very cautious to claim otherwise.

The only way to infer that the GMPs were different is to look at the temporal structure of the movement, i. The amount of time spent on a shot is not an invariant feature per se. Although, what was interesting in the MT analysis, was the fact that VP significantly lengthened the MT in the posttest. Considering the aforementioned shortcomings and results we decided to conduct another experiment, using different participants and the same task. We changed also testing conditions.

Instead of a posttest as used previously Breslin et al. As opposed to posttest, retention test assesses the learning better and gives us the idea of relative permanence or persistence of learning Magill and Anderson, We decided to use a 4-point scoring system as used previously by Breslin et al.

Twenty healthy participants were randomly divided into two groups: variable practice group VP1 — number one differs this group from the group in Experiment 1 and CP group CP1. Five participants in VP1 trained rugby, one athletics, one hockey, one cricket, and one netball.

In CP1, four participants trained rugby, one athletics, one tennis, one netball, one cricket, and one squash. One participant in CP1 and one participant in VP1 reported no previous experience in any sport. Over five consecutive days in total, the first and the 5th day were dedicated to the pretest and retention tests, respectively, whereas days 2—4 3 days in total were designated to the acquisition phase [similarly as in Breslin et al.

Unlike in Experiment 1, and like in Breslin et al. The shot efficiency was calculated based on a 4-point scoring system used in previous studies Hardy and Parfitt, ; Keetch et al. The percentage of the shot efficiency was calculated out of 60 points, i. During the acquisition phase, CP1 participants performed shots per day per one distance — free-throw distance 4. The VP1 participants performed shots per day per five distances 3.

All participants accumulated shots in the acquisition phase days 2—4 and shots in pretest and retention tests. The informed consent forms and personal information questionnaires were collected before the testing commenced.

The participant then had a five to min warm-up period designated to stretching and preparing for the experiment.

Shots were taken in a quasi-random order, with no more than two shots per distance in a row. Participants were shown how to execute shots using two hands from above their heads. The demonstration was done by the same person, the main investigator, who was familiar with basketball. Unlike in Experiment 1, tests and practice sessions took place in a laboratory where cameras and professional basketball board were mounted to the wall. All shots were recorded with a digital high-speed camera Exilim, Casio EX-ZR10, 40 fps in order to score shots appropriately.

The first five shots taken from the free-throw line i. We defined starting and ending points similarly to the Lam et al. Kinematic data analysis was done using the same procedure as in Experiment 1. Additionally, we analyzed inter-joint coordination for each throw for the shoulder—elbow and elbow—wrist joint pairs in the anterior-posterior y and vertical z axes using correlation coefficients De Oliveira et al.

In order to detect especial skill, we applied the method used originally by Breslin et al. Average percentage accuracy scores based on a 4-point scoring system were calculated for all the trials across each of the five distances. The mean values for each kinematic landmark are presented in Table 6. Table 5. Results of the linear mixed model for Experiment 2: F -values, p -values, and ES — effect sizes. Main effects group: CP1 vs. VP1; test: pretest vs. Table 6. Results of the pairwise comparisons for landmarks A—E.

Mean values of relative timing for constant CP1 and variable VP1 practice groups, p -values p and effect sizes ES provided in columns comparison between CP1 and VP1 in pre- and retention test, accordingly and rows comparison between pre- and retention test results for CP1 and VP1, respectively. We found significant test effect pre- vs. We also found group significant effect CP1 vs. VP1 in landmarks B and D. We run additional pairwise comparisons for significant effects see Table 6.

We found that in most cases the interaction effect was due to the significant differences between CP1 and VP1 groups in the pretest for all landmarks as well as due to the significant differences between pretest-retention test results in CP1 for all landmarks. We did not find any significant differences between CP1 and VP1 in the retention test. The mean values of relative timing for all landmarks in pre- and retention test in CP1 and VP1 are visualized in Figure 2. As it can be noticed, the relative timing in CP1 and VP1 groups differed significantly in the pretest, but in the retention test, both groups had, noticeably, similar results.

On the other hand, in Figure 3 , we showed the differences percent points between landmarks for CP1 and VP1 in pre- and retention test. The biggest difference was noticed in landmarks B and C in CP1 group, and it is reflected in big effect sizes see Table 6. Figure 3. Differences in relative movement time between pre- and retention test results measured for CP1 and VP1 group.

Percent points are used as measure units. The solid lines represent mean for all landmarks differences in relative timing between pre- and retention test for both groups. Similarly to the analysis in Experiment 1, we looked at the movement time.

We used the same linear mixed model to find any differences between CP1 and VP1 in pre- and retention test Table 7. Table 7. VP1; test: pre- vs. Although, there were no significant differences between MT in pre- and retention test for CP1 group, similar trend as in Experiment 1 was found — the MT shortened in retention time as compared to the pretest. Again, similarly to the results in Experiment 1, MT in retention test in VP1 was longer in the retention test as compared to the pretest.

This difference was significant and the effect size was moderate 0. Table 8. Results of the pairwise comparisons for MT. In order to determine whether coordination in CP1 and VP1 was characterized by a different pattern, we analyzed correlation coefficients for shoulder-elbow and elbow-wrist in the z and y -axis see Table 9 for coefficients values. We examined the inter-joint coordination applying the same linear mixed model as previously but we did not find any significant main effects.

Table 9. It can be noticed that correlations were moderate elbow — wrist Y and strong shoulder — elbow and elbow — wrist X. The especial skills effect was not present in either group, in pretest or retention test. Due to the fact, that predicted shot efficiencies at the 4. In the VP1 group, the real shot efficiency at 4. Our two objectives were the same as in Experiment 1, i. Secondly, we wanted to determine whether limited practice in constant conditions leads to the development of an especial skill.

However, we changed testing and practice conditions as compared to Experiment 1. We used a 1-day retention test to detect especial skill in our participants and a 4-point scoring system Hardy and Parfitt, ; Keetch et al.

We also changed the practice conditions. Instead of 5 days of practice, our participants practiced for 3 days only, they were tested at five distances [as in Breslin et al. Although we used a different scoring system, we did not detect an especial skill in CP1. It is difficult to say why we were unsuccessful. Perhaps, due to the differences in characteristics of the participants used in our experiments, or due to the previous experience participants had had, the amount of practice our participants received was too little to develop an especial skill.

We cannot claim that using a different method of detecting especial skills, e. The significant interaction effect was mostly due to the significant differences between CP1 and VP1 in the retention test as well as the significantly longer MT in the retention test as compared to the pretest in VP1. The inter-joint correlation was strong and moderate and except for one correlation shoulder — elbow in axis Z they did not differ one from another.

It means that the movements were quite consistent in both groups in pre- and retention tests. It can be noticed in relative timing analysis that we probably had recruited two different groups of participants. However, what is even more interesting, there were no differences between CP1 and VP1 in retention testes.

It means that our two groups with different or no movement patterns at the beginning were developing the same GMPs, although they practiced in different practice conditions.

The primary objective of this study was to determine whether practice in constant conditions leads to the development of different GMP as opposed to the variable practice conditions. The secondary objective was to determine whether limited practice in constant conditions leads to the development of especial skill, i. We found that a skill developed in CP and VP have the same relative timing in retention test Experiment 2. The difference was mostly due to the differences found between pretest-posttest results.

The effect of group CP vs. VP results was not significant. Considering our findings in both experiments, specifically, results obtained in Experiment 2, in which two groups CP1 vs. VP1 significantly differed one from another in pretest but they did not in retention test, we may claim that regardless of the practice conditions, variable and CP develop same GMP. We did not succeed at recreating an especial skill effect in either of our experiments. It was particularly surprising in Experiment 1, in which our participants received a bigger amount of practice shots accumulated during practice sessions as compared to that received by Breslin et al.

We may speculate, that participants in Breslin et al. Of course, it could be otherwise — participants in our experiments may have had more experience in basketball itself or in a task similar to the basketball free-throw shooting. Whatever is true, we may assume that our participants differed from experienced ones. In another study by Breslin et al. Breslin and colleagues found that correlation coefficients at the distance 4.

These differences were even higher when comparing anterior-posterior direction y -axis. In Breslin et al. In our participants see Table 9 these correlations were medium above 0. This phase is called freezing Savelsbergh et al.

On the other hand, Breslin et al. Of course, there may have been many other reasons why we did not recreate an especial skill effect in our participants. However, we believe that the failure in the replication of Breslin et al. We consider our findings regarding changes and differences in movement time as very interesting.

In Experiment 1, participants practicing in variable practice had significantly longer movement time than CP participants. A similar trend was found in Experiment 2. In previous studies on especial skill Keetch et al. Practice in constant conditions may facilitate parameter assignment Keetch et al.