This article was downloaded by: [Fordham University]

On: 04 August 2013, At: 08:44

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Educational Research
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rere20

The relative age effect and the

development of self-esteem
a b c
Angus H. Thompson , Roger H. Barnsley & James Battle
a
University of Alberta, Canada
b
The University College of the Cariboo, Canada
c
Consulting Services, Edmonton Public Schools, Canada
Published online: 10 Dec 2010.

To cite this article: Angus H. Thompson , Roger H. Barnsley & James Battle (2004) The relative
age effect and the development of self-esteem, Educational Research, 46:3, 313-320, DOI:
10.1080/0013188042000277368

To link to this article: http://dx.doi.org/10.1080/0013188042000277368

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
 Educational Research, Vol. 46, No. 3, Winter 2004

The relative age effect and the development

of self-esteem

Angus H. Thompson1*, Roger H. Barnsley2 and James Battle3

1
University of Alberta, Canada; 2The University College of the Cariboo, Canada;
3
Consulting Services, Edmonton Public Schools, Canada

A recent paper has demonstrated a relationship between suicide during the teen years and the age,
relative to one’s classmates, at which these individuals entered school. This represents the latest,
Downloaded by [Fordham University] at 08:44 04 August 2013

and perhaps most important, of a series of studies that have focused on the effects of grouping
children by age of entry into particular activities. This phenomenon, known as the relative age
effect, is strikingly evident in activities that are competitive and where performance is highly
correlated with age and level of maturity. To date, relative age research has reported significant and
substantial achievement differences within the confines of athletic and academic pursuits. However,
with the advent of the study noted above, it now appears that emotional development is also
implicated. Here we demonstrate that a relatively young age of entry into the formal educational
system is associated with reduced self-esteem several years later. This suggests that self-esteem (or a
related factor, such as self-efficacy) serves as an important factor lying functionally between
proximal relative age effects and suicide.

Keywords: Emotional development; Family structure; Relative age; Self-esteem; Suicide

Introduction
The relative age effect in sport was first noted among élite-level ice hockey players
when it was observed that they tended to be born early in the eligibility year for youth
entry into the sport, which for hockey corresponds with the calendar year (Grondin et
al., 1984; Barnsley et al., 1985). These findings demonstrated that within the major
junior leagues and the National Hockey League, players’ birth dates decreased in
prevalence from January through December. It was theorized that this relative age
effect arose from the consequences of grouping young boys for entry into organized
minor hockey, thereby producing a one-year age range for the participants. As size,
speed and coordination are highly correlated with age, older players within the age-
group will, on average, show superior performance.

* Corresponding author: Department of Public Health Sciences, 13 – 103 Clinical Sciences

Building, University of Alberta, Edmonton, Alberta T6G 2G3, Canada.
Email: gus.thompson@ualberta.ca
ISSN 0013-1881 (print)/ISSN 1469-5847 (online)/04/030313-08
# 2004 NFER
DOI: 10.1080/0013188042000277368
 314 A. H. Thompson et al.

Thus, it can be said that maturity had been mistaken for ability by coaches, peers
and the individuals themselves. The resulting expectations that are created for
individual children creates a self-fulfilling prophecy that provides age-advantaged
children with greater self-confidence and regard by others. The opposite will likely
hold for those younger than the average in their age-group, with adjustments to
poorer initial performance likely, including lowered self-confidence and self-esteem.
One consequence that has been found is an increased drop-out rate for those youthful
hockey players that had been disadvantaged by age in the past (Barnsley &
Thompson, 1988), suggesting that given the choice, younger children will seek to
leave or avoid an activity in which their competitive position is hampered by their
relative age. Interestingly and predictably, the relative age effect has also been found
in other competitive sports such as baseball (Thompson et al., 1991), world-class
soccer (Barnsley et al., 1992; Dudink, 1994) and American football (Glamser &
Marciani, 1992). From this research it can be reliably concluded that the relative age
Downloaded by [Fordham University] at 08:44 04 August 2013

effect will be evident, if not profound in children’s sports programmes that are
characterized by age grouping, competition and selection by ability.
Having arrived at this conclusion, it should be noted that schooling shares many of
the structural characteristics of children’s sports programmes. Children are age
grouped for entry into school, are rated according to achievement, and are placed in
different programmes with different curricula and learning options based on these
measures of performance. In line with this logic, a large body of literature has
consistently reported a strong effect of the age of school entry on academic
achievement.
In England, Pidgeon and Dodds (1961) and Jinks (1964) noted that chronological
age could not account for such findings, thus paving the way for other explanations
such as duration of instruction and relative age. Later investigations showed that
school children with a relative age advantage are more likely to have higher
achievement (Thompson, 1971; Allen & Barnsley, 1993), to be placed in
programmes for gifted children (Maddux et al., 1981) and to be placed in more
challenging educational streams or classes (Freyman, 1965; Sutton, 1967). Children
with a relative age disadvantage are more likely to be retained (‘failed’) for an
additional year in the same school grade (Langer et al., 1984), to be referred for
psychological assessment (DiPasquale et al., 1980) and to be placed in a specialized
group or provided with a diagnostic label for remedial instruction (Maddux, 1980;
Diamond, 1983). The relative age phenomenon is apparently long-lasting, as effects
have been found among 13-year-olds (Kalk et al., 1982) and 15-year-olds (Bell &
Daniels, 1990).
As a consequence of these findings, the suggestion has arisen that parents, at least
in North America, should postpone school entry for those younger children whose
birthday places them near the ‘cutoff’ for their age-group (Uphoff & Gilmore, 1986).
The result of such action would place the children in question among the eldest of
their eventual classmates, rather than the youngest. This contradicts an earlier
tendency of parents to try to arrange early admission for children who were actually
too young to make a particular cutoff (Barnsley & Thompson, 1985). It should be
noted that such parental discretion is not available in all countries. In England, for
 Short reports 315

example, almost all children are educated in the ‘correct’ grade for their age
(Hutchison & Sharp, 1999).

The present study

The relevance of the above to the present study lies in a hypothesized link that was
proposed between relative age and suicide. The proposed causal chain involves the
following steps: first, relative age produces differences in achievement that are due to
maturation, not ability. Second, these differences lead to variation in self-esteem and
confidence. Third, low self-esteem and lack of self-confidence are associated with a
child’s inability to compete with his or her classmates leading to depression and
hopelessness respectively. Finally, depression and hopelessness, which have been
regarded as essential ingredients of suicide (Beck et al., 1975; Dyer & Kreitman,
1984), become precursors of self-harming behaviour. In fact, a study based on this
Downloaded by [Fordham University] at 08:44 04 August 2013

logic produced findings that were in line with its basic premise. That is, youth
suicides in Alberta, Canada, showed higher rates among those who were in the
younger half of the school-entry cohort than for those who formed the older portion
(Thompson et al., 1999).
This finding, which we think has profound implications for social policy, has
rekindled our interest in a study of the hypothesized intermediate role of self-esteem.
Simply put, our interpretation of the relative age findings points to lowered self-
esteem early in life as a significant link between relative age at school entry and the
likelihood of later suicidal behaviour. The purpose of this study, then, was to
evaluate the possibility that self-esteem can be determined by age of school entry. It
should be noted that this study was conducted with the awareness of the criticisms
that have been levelled at the self-esteem construct (Burr & Christensen, 1992). Of
particular importance is Seligman’s (1991) view that, while self-esteem may reflect
an important condition, changing it does not affect other behaviours that are
associated with it. With this in mind, self-esteem here will serve as an indication of
some form of self-evaluation, without any implication that its manipulation will
necessarily ameliorate any associated conditions such as depression or suicidal
behaviour.
There is some reason to expect that relative age effects may vary according to the
amount of stress that children have to deal with. Rutter (1979) has suggested that
while the presence of only one risk factor has virtually no effect, additional stressors
produce an exponential increase in emotional disturbance in children. Social
problem frequency has a similar supposed effect on the likelihood of exhibiting a
psychiatric disorder among adults (Thompson & Bland, 1995). Thus, the analysis
to be described below includes a measure of family stress (i.e. intact versus non-
intact family structure). The predictions made were (1) greater relative age at 1st
grade entry will be associated with increased self-esteem, (2) those who lived with
both parents will exhibit higher self-esteem than those who lived with neither, or
only one, parent, and (3) those from intact families may not show a relative age
effect that is as strong as those who had the additional disadvantage of a broken
home environment.
 316 A. H. Thompson et al.

Method
The database
The database used for analysis was created by matching individual student records
from two existing electronic data bases. This resulted in a data file containing records
of 1129 students, 46% in 1st to 3rd grades, 44% in 4th to 6th grades and virtually all
of the remaining 10% being in the 7th to 9th grades. The file included general
demographics and information on age at school commencement, self-esteem, and
family structure, as well as other academic data not used in the present study. Once
the matching process was completed, personal identifiers (e.g. student names, ID
numbers) were deleted from the research file.
Individuals under age 16 at time of testing were administered the child form of the
Culture Free Self-Esteem Inventory (Form A: Battle, 1981), and the remainder were
administered the adult form (Form AD). This paper-and-pencil test was standardized
Downloaded by [Fordham University] at 08:44 04 August 2013

on boys and girls in the 2nd to 9th grades in Canada and the USA. This instrument
comprises a number of sub-scales (not presented here) that can be summed to
produce the total self-esteem score that was used in the present study. Test – retest
reliabilities as reported in the Manual (Battle, 1992) ranged from 0.72 to 0.93 for
Form A, and 0.79 to 0.82 for Form AD. As suggested in the Manual, raw scores were
converted to T-scores (mean = 50, Std. Dev. = 10) to allow meaningful comparisons
across test forms.

Analysis
Relative age simply means one’s age in comparison to the age of others in that
person’s cohort. Here, we used the calculation of age (in years and months) at 1st
grade entry, regardless of ‘current’ age and grade. In accord with regular school
procedure, those with the correct age for 1st grade entry would range from age 5 years
6 months to 6 years 5 months. We also analysed the records of those falling within
three months of the outer limits of the correct age for grade limits to allow an
examination of the data for those who entered 1st grade at an extraordinary age. More
extreme age departures were not considered in the analysis.

Results
Mean total self-esteem data according to age at entry into 1st grade (grouped into
three-month age categories) and family structure (living with both parents versus one
or neither) are shown in Figure 1. In general, we see a linear increase in self-esteem as
age at entry increases (F1,1105 = 2.49, p 5 0.03), with a post hoc analysis showing no
significant departure from linearity (F4,1123 = 0.17, p 5 0.96). This was true
regardless of family structure, but the self-esteem of those from broken homes was
considerably lower at all of these ages (F1,1105 = 20.36, p 5 0.001). The effect of
family life on child development has often been mentioned (e.g. Bronfenbrenner,
1979), and is corroborated here by this finding of a substantial family structure effect.
However, we did not find an interaction between relative age and family structure
 Short reports 317
Downloaded by [Fordham University] at 08:44 04 August 2013

Figure 1. Variation in total self-esteem as a function of family structure and age at entering 1st grade

(F5,1105 = 0.40, p 5 0.85), thus providing no support for our hypothesis that life
within an intact family would neutralize relative age effects. Family structure and
relative age, then, appear to operate independently and in an additive manner.
It is interesting that the highest self-esteem was displayed by those who were over-
age for 1st grade entry. Why did these children enter 1st grade a year later than their
eligibility allowed? C. Varnhagen, of the University of Alberta (personal commu-
nication, 1992), has noted that the school board in question allows parents of
relatively young, 1st grade eligible children, to hold back these youngsters for one year
(about 20% to 30% take advantage of this offer), in effect, changing the status of these
children from being among the youngest in the class to being counted among the
eldest when they do eventually begin the 1st grade.

Discussion
Educators, and many of us who have reminisced about our early influences, will
intuitively grasp the causal relationship between competition with others who are
older than ourselves and subsequent self-assessment and emotional response. Yet, to
our knowledge, this is the first time that this has been demonstrated empirically. Our
findings implicate structural aspects of society that may have been set up for one
 318 A. H. Thompson et al.

purpose, but may produce harm to some that ‘out-balance’ the intended positive
effects.
This is particularly notable in relation to differences observed between the early
and late entrants. Given that the early entrants were probably perceived by parents,
school district and teachers as being precocious and thus having the ability to
function, if not excel, in school at an early age, it is important to note that a
consequence of this decision was the lessening of the self-esteem of these individuals.
As ability and self-esteem interact in direct and subtle ways to influence long-term
achievement, risk-taking and, ultimately, life successes, it is paradoxical, and should
be of great concern, that these potentially highly achieving children are perhaps being
impeded, rather than advantaged. In fact, the data suggest that had these children
enrolled in school at the regular time, their self-esteem would have been substantially
higher, thereby producing the opportunity for the interaction with their exceptional
abilities that was originally sought.
Downloaded by [Fordham University] at 08:44 04 August 2013

Similarly, the findings with regard to the late entrants are really counter-intuitive.
The late entrants, some of whom may have been judged to require extra time for
maturation, realized the highest self-esteem of all. Perhaps the overriding motivation
in holding these children back was not to deal with a deficit in maturation, but rather
to take advantage of the relative age benefits that accrue to those who are held back.
This is a matter for further research. In any case, to the extent that high self-esteem
may mitigate developmental disadvantage, the decision to delay school entrance was
well considered.
This finding, then, has particular relevance for parents who may be faced with
questions regarding their child’s entry into school or other competitive, age-grouped
activities and, especially, for those who may be considering an attempt to enter their
child early. The evidence is quite strong that such children are at a substantial
disadvantage and that the long-term effects are potentially damaging. The attractive
action, in this context, is to simply hold a relatively young child out of school for one
year. This would allow an additional 12 months of maturation before placement in
the competitive classroom environment, thus putting the child in question in a
position of relative age advantage with an attendant shift in the odds towards healthier
emotional development and greater achievement. While this is fine from the view of a
parent, from a public policy point of view it should not be expected to provide a
benefit to children overall. Each child who gains advantage by being held back
exposes another at the other end of the spectrum to a matched level of disadvantage.
Clearly it is unacceptable that such important advantages and disadvantages accrue to
children simply by virtue of their birth date.
While it is clear that relative age is associated with self-esteem, it should be
remembered that this investigation has dealt only with relative age effects due to age
of school entry. There will also be similar effects due to age of entry into various
sports and other competitive pursuits. Furthermore, other structural aspects of our
society that confer advantage or disadvantage would also affect the self-appraisal of a
developing child in a similar fashion. Indeed the amount of variance accounted for
here is not large. Future investigations may benefit from consideration of relative age
effects in a more comprehensive manner.
 Short reports 319

From a societal point of view, it should be noted that relative age, while of great
structural importance, is not the defining concept of its own effect. We see it as one
aspect of a larger notion, the development of self-appraisal and the conclusions drawn
from it. This involves a number of other related, and sometimes competing, concepts
like self-efficacy, self-esteem, mastery, self-image, and so on. It is not our purpose to
deal with this matter here, but to point out that we are contending with an entity that
is much broader than birth date, and that in some form, has a significant effect on
later happiness, health and social behaviour. It should be a goal of further research to
identify ways in which the mechanisms that produce relative age differentials that
favour some children over others can be changed, so that the benefits of any activity in
question can be optimally applied to all children. But this structural approach does
not represent the whole story. There are many ways to ensure that all children
develop positive self-appraisals, and many of these will buffer negative relative age
effects or supplement the positive ones. Research is needed to account for these and
Downloaded by [Fordham University] at 08:44 04 August 2013

to describe their function in this context.

Acknowledgements
Funding for this study was provided by the Muttart Foundation and by the Alberta
Centre for Injury Control and Research of the University of Alberta, both of
Edmonton, Canada. We also acknowledge the indispensable contributions of Delia
Kuzz, Bruce McIntosh, and Dale Armstrong of the Edmonton Public Schools, and
Trustee Don Fleming of the Edmonton Public School Board.

References
Allen, J. & Barnsley, R. H. (1993) Streams and tiers: the interaction of ability, maturity and training
in systems with age-dependent recursive selection, Journal of Human Resources, 28, 649 – 659.
Barnsley, R. H. & Thompson, A. H. (1985) Gifted or learning disabled: the age of entering school
may make the difference, Early Childhood Education Council Journal, 18, 11 – 14.
Barnsley, R. H. & Thompson, A. H. (1988) Birthdate and success in minor hockey: the key to the
NHL, Canadian Journal of Behavioural Science, 20, 167 – 176.
Barnsley, R. H., Thompson, A. H. & Barnsley, P. E. (1985) Hockey success and birth-date: the
relative age effect, Journal of the Canadian Association for Health, Physical Education and
Recreation, 51, 23 – 28.
Barnsley, R. H., Thompson, A. H. & LeGault, P. (1992) Family planning football style: the relative
age effect in football, International Review for the Sociology of Sport, 27, 77 – 78.
Battle, J. (1981) Culture-free self-esteem inventory for children and adults. (Seattle, WA, Special Child
Publications).
Battle, J. (1992) Culture-free self-esteem inventories (2nd edn) (Austin, TX, Pro-Ed).
Beck, A. T., Kovacs, M. & Weissman, M. M. (1975) Hopelessness and suicidal behaviour: an
overview, JAMA, 234, 1146 – 1149.
Bell, J. F. & Daniels, S. (1990) Are summer-born children disadvantaged? The birthdate effect in
education, Oxford Review of Education, 16, 67 – 80.
Bronfenbrenner, U. (1979) The ecology of human development: experiments by nature and design
(Cambridge, MA, Harvard University Press).
Burr, W. R. & Christensen, C. (1992) Undesirable side effects of enhancing self-esteem, Family
Relations, 41, 460 – 464.
 320 A. H. Thompson et al.

Diamond, G. H. (1983) The birthdate effect? A maturational effect? Journal of Learning Disabilities,
16, 161 – 164.
Dipasquale, G. W., Moule, A. D. & Flewelling, R. W. (1980) The birthdate effect, Journal of
Learning Disabilities, 13, 4 – 8.
Dudink, A. (1994) Birth date and sporting success, Nature, 368, 592.
Dyer, J. & Kreitman, N. (1984) Hopelessness, depression and suicidal intent in parasuicide, British
Journal of Psychiatry, 144, 127 – 133.
Freyman, R. (1965) Further evidence on the effect of date of birth on subsequent school
performance, Educational Research, 8, 58 – 64.
Glamser, F. D. & Marciani, L. M. (1992) The birthdate effect and college athletic participation:
some comparisons, Journal of Sport Behavior, 15, 227 – 238.
Grondin, S., Deshaies, P. & Nault, L. P. (1984) Trimestres de naissance et participation au hockey
et au volleyball, La Revue Québecoise de l’Activité Physique, 2, 97 – 103.
Hutchison, D. & Sharp, C. (1999) A lasting legacy? The persistence of season of birth effects, paper
presented at the Conference of the British Educational Research Association, Brighton, 2 – 5
September.
Downloaded by [Fordham University] at 08:44 04 August 2013

Jinkes, P. C. (1964) An investigation into the effect of date of birth on subsequent school
performance, Educational Research, 6, 220 – 225.
Kalk, J. M., Langer, P. & Searls, D. T. (1982) A closer look at school cutoff dates and achievement
(Report No. AY-AA-52) (Denver, CO, National Assessment of Educational Progress,
Educational Commission of the States), February.
Langer, P., Kalk, J. M. & Searls, D. T. (1984) Age of admission and trends in achievement: a
comparison of blacks and Caucasians, American Educational Research Journal, 21, 61 – 78.
Maddux, C. D. (1980) First-grade entry age in a sample of children labelled learning disabled,
Learning Disabilities Quarterly, 3, 79 – 83.
Maddux, C. D., Stacy, D. & Scott, M. (1981) School entry age in a group of gifted children, Gifted
Children Quarterly, 15, 180 – 184.
Pidgeon, D. A. & Dodds, E. M. (1961) Length of schooling and its effect on performance in the
junior school, Educational Research, 3, 214 – 221.
Rutter, M. (1979) Protective factors in children’s responses to stress and disadvantage, in: M. W.
Kent & J. E. Rolf (Eds) Primary prevention and psychopathology. Vol. III, Social competence in
children (Hanover, NH, University Press of New England).
Seligman, M. E. P. (1991) Learned optimism (New York, A.A. Knopf).
Sutton, P. (1967) Correlation between streaming and season of birth in secondary schools, British
Journal of Educational Psychology, 37, 300 – 304.
Thompson, D. (1971) Season of birth and success in the secondary school, Educational Research,
14, 56 – 60.
Thompson, A. H., Barnsley, R. H. & Dyck, R. J. (1999) A new factor in youth suicide: the relative
age effect, Canadian Journal of Psychiatry, 44, 82 – 85.
Thompson, A. H., Barnsley, R. H. & Stebelsky, G. (1991) Born to play ball: the relative age effect
and major league baseball, Sociology of Sport Journal, 8, 146 – 151.
Thompson, A. H. & Bland, R. C. (1995) Social dysfunction and mental illness in a community
sample, Canadian Journal of Psychiatry, 40, 1 – 6.
Uphoff, J. & Gilmore, J. (1986) The birthdate effect and college athletic participation: some
comparisons, Journal of Sport Behaviour, 15, 227 – 238.