Download presentation
Presentation is loading. Please wait.
1
Dale Ulrich, University of Michigan
Youngsik Park and Weimo Zhu, University of Illinois at Urbana-Champaign Dale Ulrich, University of Michigan Finding Unexpected Response Patterns in TGMD-2 Using Social Network Analysis ILLINOIS KINESMETRICS BACKGROUND/NEED RESULTS Agreement of clustered groups between the two datasets Contingency coefficient = .787, Kappa = .522 Distribution of ages by the clustered groups Overall, ages are related to clustered groups. Between two different occasion data, there were no age difference among the clustered groups. Characteristics of clustered groups According to ability level Three to four groups could be found, high, (mid-high), low, (very low). According to subtests (motor and objective skill test) Two motor skill oriented groups were found. The one (group 2) from high correct score groups and the other (group 6) from low correct score groups. One objective skill oriented group was found. Group 5 showed higher scores in objective skills than group 4 which showed higher group mean score than group 5. Group 3 indicated a little motor dominant response pattern and group 1 indicated a little objective dominant response pattern. Group 4 did not show any explicit response pattern in the subtest level. According to 12 skill tests and 48 subskill items Group 2 and 6 Group 1 and 2 showed overall high ability level across all skill tests. A major difference between two groups was found in the response patterns for overhand throw (objective) skill test. The correct score of group 1 is .867 and that of group 2 is .35. Similar answering patterns were detected in all four subskill items of overhand throw skill test. Another difference between two groups was that group 1 showed much higher mean correct scores in three most difficult items among motor skill test. Group 3 In most of skill tests, this group was around the mid-high range in their ability. Much lower correct scores than expected were detected in two difficult items - one from gallop skill test and one from strike skill test. Group 4 A little lower mean scores were indicated in two skill tests such as jump and dribble and their subskill items. TGMD – 2 ( Ulrich, 2000) is a well-known standardized test to assess gross motor skills for children age 3 to 11. The test consists of 12 skills – six skills each for two subtests, locomotor and object control. It is assumed that students who have similar summary scores would respond in a similar manner across all items. In practice, participants do not follow expected patterns. In aptitude assessment, those unexpected patterns have been used to identify invalid response patterns, such as cheating or guessing (Meijer & Sijtsma, 1995, 2001), and in personality domain, they were used for broader purposes such as detecting social desirability (Zickar & Drasgow, 1996) or person-fluctuation (Ferrando, 2004). These unexpected patterns, which may provide rich information for instruction or intervention, have been basically ignored in motor assessment. Mean Scores Item Clustered Groups Item name Item code 1 2 3 4 5 6 7 Total Run item.1.1.1 0.66 0.96 0.84 0.80 0.51 0.98 0.77 0.86 item.1.1.2 1.00 0.92 0.94 item.1.1.3 0.87 0.89 item.1.1.4 0.91 0.82 0.28 0.95 0.70 0.85 Gallop item.1.2.5 0.32 0.62 0.50 0.30 0.04 0.27 0.47 item.1.2.6 0.59 0.90 0.68 0.13 0.93 0.63 item.1.2.7 0.75 0.43 0.72 0.88 item.1.2.8 0.48 0.64 0.09 0.52 0.78 Hop item.1.3.9 0.36 0.39 0.00 0.33 item 0.60 0.06 0.73 item 0.38 0.55 item 0.61 item 0.45 0.34 0.79 Leap item 0.74 0.15 0.67 item 0.76 0.83 item 0.58 0.25 0.42 Horizontal jump item 0.81 0.41 item 0.44 0.19 item item Slide item 0.56 0.71 item 0.97 item item 0.57 0.11 Striking a stationary ball item item 0.65 0.54 0.69 item item 0.35 item Stationary dribble item 0.02 item 0.29 0.20 item item Catch item item 0.26 item 0.17 Kick item 0.49 item item item Overhand throw item 0.46 item 0.22 item item Roll item item 0.31 0.53 item item PURPOSE In this study, the children who showed unexpected response patterns in TGMD-2 were detected and the characteristics of the patterns were described using social network analysis approach. METHOD About Participants Number of participants: N = 1143 Data were collected two times (occasions) from the same subjects. Male (571, 50%), female (572, 50%), age 3 to 10 years old About Items Number of items was 48 and the test consists of two subtests – locomotor and object control. Each subtest consists of six skill tests and each skill test have three to five subskill items. Analysis Procedure The raw data were pre-processed using the social network approach. After pre-processing , the data were fitted to regression lines Finding clustered groups After fitting, the dissimilarity measures were calculated using the residuals and clustered groups were found using dissimilarity measures. Explaining the clustered groups Using clustered groups, descriptive statistics (mean and standard deviation) for each item were calculated. Cross Validation For cross validation, the two occasional datasets were processed in the same way. The agreement of the classification by the clustered groups was calculated between the two datasets. CONCLUSIONS Frequency of the clustered groups Occasion 1: 398, 44, 129, 210, 44, 64, 53 Occasion 2: 412, 57, 186, 119, 37, 66, 65 Group 3 and 4 was divided well in each data set. But the half of the members of the two groups are classified into the other group in different datasets. (group 3 group 4 or group 4 group 3). Social network analysis is a useful approach in identifying unexpected response patterns in motor assessment, which in turn can be used to help design appropriate instruction and assessment.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.