'Training Effect' in the Design of a Longitudinal Study
| Autor: | Mirjana Rajic, Claude Beaucage, Guy R. Brisson, Jean-Claude Jéquier, Robert La Barre, Hugues Lavallée |
|---|---|
| Rok vydání: | 1974 |
| Předmět: | |
| Zdroj: | Perceptual and Motor Skills. 39:693-694 |
| ISSN: | 1558-688X 0031-5125 |
| DOI: | 10.2466/pms.1974.39.2.693 |
| Popis: | Summary.-An attempt was made to detect possible effects of familiarization ("training effect") in the interpretation of data collected repeatedly on children. During a longitudinal study on child growth we looked for such effects on different performance tests done at various ages. We noticed that work capacity tests (PW670; VO? max) were occasionally subjected to interferences from familiarization. We conclude that a proper interpretation of repeated measurements must take into account possible biases originating from "training ef fecc." By definition a longitudinal study implies the measurement of characteristics at different points in rime (MacMahon & Pugh, 1970). Some of the characteristics frequently studied are in fact performances (vital capacity, maximal oxygen consumption, projection of bodily frontiers, verticality) requiring a certain quality of participation. The difference berween performances measured at two points in time originate from an actual improvement in performance and, possibly, from an effect of familiarization with testing procedures and/or apparatus by Ss. The understanding of the familiarization effect, that we term "training effect," shall lead to a more accurate inrerpretation of results and to the determination of optimal intervals berween measurements. Thus, it was decided to analyze data obtained from a multidisciplinary longitudinal study on child growth conducted over the last 3 yr. by the Health Sciences Research Center of the University of Quebec in Trois-Rivieres. In that research are enrolled 600 elementary school children chat entered the study at various ages and that are being examined once a year on their birthmonths. Some 250 parameters are measured mainly in the fields of anthropometry, spirometry, psychology, dynamometry and work physiology. We selected six variables on the grounds of repeated subjective observations of children's reactions when first confronted with the testing apparatus. With these six variables, performances for a same age were compared between Ss being evaluated for a first, a second and, occasionally, for a third time. They were: PWC170 (Physical Work Capacity measured at 170 beats/min. on a bicycle ergometer; Howell & Macnab, 1968) expressed in Kpm/min. and in Kpm/min./kg, and VO~ max (maximal oxygen consumption) expressed in liters/min. and in ml/kg/min. For these four measures repeated data were available for Ss at ages 6, 7, 8, 10 and 11 yr. Two other variables, the Force Index (Clarke, 1967) and the PFI (Physical Pitness Index; Clarke, 1967) were similarly treated for age group 11 only. Statistical analyses used Student's t test, except for those distributions where the hypothesis of normality seemed violated. For the latter, the non-parametric tests of Mann-Whitney and of Smirnov (Conover, 1971, pp. 224, 309) were used. Out of the six studied variables, three were not significantly affected by "training," i.e., did not change with repetition of testings whatever the age considered (PWCXO in Kpm/min./kg; Force Index; PFI). For the three other variables, significant differences occurred only at certain ages. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|