THE RELATIVE AGE EFFECT IN WOMEN’S ICE HOCKEY: INTERNATIONAL AND POSITIONAL COMPARISONS

A relative age effect (RAE), or chronological age differences among individuals within the same age group, has been observed in 14 sports (Cobley et al., 2009). A strong RAE has been reported in men’s ice hockey since 1985 (Barnsley et al., 1985). In contrast, research on RAE in women’s ice hockey is limited (Wattie et al., 2007; Weir et al., 2010) and findings are equivocal.PURPOSE: To determine whether there is a RAE in women’s ice hockey, and if it varies by country or player position.METHODS: Participants were 117 female ice hockey players (mean age=19.9±2.3 yr) on 2 Swedish elite-level club teams (n=54) and 4 Canadian university teams (n=63). Players reported birthdate and position (Forward=F, Defenseman=D, Goalie=G). Birthdates were coded by quartiles (Q1: Jan.-March, Q2: April-June, Q3: July-Sept., Q4: Oct.-Dec.) and by half-year. Birthdate data were submitted to chi-square analyses for the sample, by country, and by position. SPSS 17.0 for Windows was used for all analyses (significance level: p<0.05). RESULTS: A significant RAE was observed for the entire sample by quartile (Q1: 28.2%, Q2: 34.2%, Q3: 25.6%, Q4: 12.0%; χ2=12.402, p=0.006). More players were born in the first half of the year than the second (62.4% vs. 37.6%, respectively; χ2=7.188, p=0.007). In contrast, Q4 was underrepresented for the sample and by country (Canada: 12.7%, Sweden: 11.1%). A RAE was present for the Canadian players by quartile (χ2=13.381, p=0.004) and half-year (χ2=9.921, p=0.002); however, no RAE was observed for Swedish players. In addition, a significant RAE was observed for the entire sample by position for G and D by quartiles (G - χ2=10.077, p=0.018; D - χ2=8.444, p=0.038) and half year (G - χ2=6.231, p=0.013; D - χ2=4.000, p=0.046), but not for F.CONCLUSIONS: The significant RAE in this sample is consistent with that in men’s ice hockey and the tendency for greater participation by relatively older players in women’s ice hockey. RAE absence in the Swedish players may reflect lower participant number, competitive level, and sociocultural support, as well as greater variation in skill level. The significant RAEs observed in Canadian players and by position support the findings of Weir et al. (2010), but the positional differences found were inconsistent, perhaps due to differences in sample size and competitive level between studies.