52 Faculty of Sport, University of Ljubljana, ISSN 1318-2269  Kinesiologia Slovenica, 17, 1, 52–62 (2011)
IZVLEČEK
Ni veliko raziskav, ki bi se ukvarjale s tipičnimi 
spretnostmi in motoričnim znanjem v znanosti o športu 
in vadbi ter na ta način preučevale učinke razvrščanja po 
skupinah. Cilj pričujoče raziskave je bil preučiti različne 
učinke razvrščanja po sposobnostih na izboljšanje 
izbranih plavalnih sposobnosti v primerjavi s prakso 
razvrščanja v heterogene skupine. Vzorec merjencev je 
obsegal univerzitetne študente (moški, stari 21 ± 0,9 let 
in v dobrem zdravstvenem stanju), ki so bili razdeljeni 
na eksperimentalno (E; n = 37) in kontrolno skupino (K; 
n = 42). Obe skupini sta bili vključeni v devetmesečni 
plavalni študijski program in sta v raziskavi sodelovali 
trikrat na teden. Skupina K je bila naključno razdeljena 
v tri heterogene podskupine, skupina E pa v tri 
homogene podskupine (neuspešni, povprečni in zelo 
uspešni). Vzorec spremenljivk je vseboval dvanajst 
spremenljivk sposobnosti plavanja, ki so bile izmerjene 
pred (I) in po izvedbi preučevanega programa (F), ter 
mnenja študentov o njihovi praksi učenja. Analiza 
variance je pokazala koristne učinke razvrščanja 
glede na sposobnosti pri zelo uspešnih in povprečnih 
plavalcih, pri neuspešnih pa med programom K in E ni 
bilo značilnih razlik. Na podlagi rezultatov raziskave 
ni mogoče potrditi predhodnih negativnih ugotovitev 
glede razvrščanja po sposobnostih, kjer so avtorji 
izjavili, da lahko ta praksa pri neuspešnih študentih 
spodbudi nizka pričakovanja glede lastne uspešnosti. 
Ključne besede: plavanje, motorično učenje, uspešnost, 
vprašalnik, sposobnosti
*Corresponding author: 
University of Split, Faculty of Kinesiology, Nataša Zenić, PhD
Teslina 6, CRO-21000 Split, Croatia
E-mail: natasazenic@yahoo.com
ABSTR ACT
There is a clear lack of studies dealing with the 
characteristic skills and motor knowledge in sport 
and exercise science and studying the effects of ability 
grouping on them. The aim of the present study was 
to examine the differential effects of ability grouping 
versus heterogeneous grouping practice on improvement 
in selected swimming skills. The sample of subjects 
consisted of male university students (age: 21 ± 0.9 
years; in good health), divided into an experimental 
group (E; n = 37) and a control group (C; n = 42). 
Both groups participated in the nine-month, three-
times-a-week swimming curriculum. The C group was 
randomly divided into three heterogeneous subgroups, 
while the E group was clustered in three homogenous 
subgroups (low, average and high achievers). The sample 
of variables consisted of 12 swimming skill variables 
measured before (I) and after the studied programme 
(F) and students’ thoughts regarding their teaching 
practice. An analysis of variance showed the beneficial 
effects of ability grouping for the high and average 
achievers and no significant differences between the C 
and E programmes for the low achievers. The results 
of our study do not allow us to support some previous 
negative findings concerning ability grouping where 
it was stated that this practice may communicate self-
fulfilling low expectations in low-ability students. 
Key words: swimming, motor learning, effectiveness, 
questionnaire, abilities
1 
University of Split, Faculty of Kinesiology, Teslina 6, 
21000 Split, CROATIA
2 
University of Rijeka, Faculty of Teacher Education, 
Slavka Krautzeka bb, 51000 Rijeka
ABILITY GROUPING AND IMPROVEMENT 
IN SWIMMING SKILLS: AN ANALYSIS OF 
THE OBJECTIVE EFFECTS AND STUDENTS’ 
SUBJECTIVE ATTITUDES
RAZVRŠČANJE PO SPOSOBNOSTIH IN 
IZBOLJŠANJE PLA V ALNIH SPOSOBNOSTI; 
ANALIZA OBJEKTIVNIH UČINKOV IN 
SUBJEKTIVNEGA ODNOSA ŠTUDENTOV
Natasa Zenić
1
*
 
BiljanaTrajkovski 
2 
Mario Tomljanović
1

Ability grouping and improvement in swimming skills 53 Kinesiologia Slovenica, 17, 1, 52–62 (2011)
INTRODUCTION
In sport and exercise science (SES), various methods of improvement in teaching effectiveness 
have been suggested and studied (Fredenburg, Lee, & Solmon, 2001; Harrison et al., 1999; Sa-
riscsany & Pettigrew, 1997; Wang et al., 2005; Wilkinson, 1992). Silverman and Skonie (1997) 
indicated that most of the research published in physical education (PE; they reviewed more than 
2,700 papers) were focused on the effectiveness of the programmes. This is logical because each 
course in SES can be observed as a teaching process, irrespective of whether it is sport training, 
PE and/or a sport/recreation session. Therefore, in all disciplines of the applied SES the aim and 
one of teachers’ focuses should be improving their pedagogical (teaching, training) effectiveness 
(Škof, Cecić Erpič, Boben, & Zabukovec, 2004).
One of the approaches used for increasing teaching (training, education) efficacy is ability group-
ing (AG). Ability grouping, or tracking, is the practice of separating students into achievement 
groups and tailoring their curriculum accordingly. In theory, AG increases student achievement 
by reducing the disparity in student ability levels, and this increases the likelihood that teachers 
can provide instruction that is neither too easy nor too hard for most students. The assumption 
is that AG allows a teacher-instructor (1) to increase the pace and raise the level of instruction 
for high achievers and (2) to provide more individual attention, repetition and review for low 
achievers (adapted from Bunton et al., 2000, and Chambers, 1988). Proponents of AG argue 
that a unique, generalised curriculum short-changes both high-achieving and low-achieving 
students. They point out the advantages it offers gifted students who may not thrive unless they 
are challenged, as well as slower learners who may grow weary of trying and failing to keep up 
with their more able peers (Linchevski & Kutscher, 1998; Lou et al., 1996). By contrast, one of 
the main arguments against AG is that the practice creates classes or groups of low achievers 
who are deprived of the example and stimulation provided by high achievers. Further, there is 
some evidence that groups with a low performance often receive lower quality instruction than 
other groups. Slavin (1990 and 1991) sees the creation of “academic elites”, a practice which goes 
against democratic ideals, as the most compelling argument against AG. 
Finally, in PE and sport there is an obvious deficiency of experimental evidence of the effects of 
AG. In one of the rare studies dealing with AG effects, Zenić and Grčić Zubčević (2006) noted 
the positive effects of AG in a swimming curriculum but also pointed out the need to investigate 
the students’ attitudes to such a practice, as previously suggested by Linchevski and Kutscher, 
Lou et al., and Slavin (see above). 
Zenić and Grčić Zubčević (2006) discussed the reasons for the lack of experimental studies on 
AG in SES and identified the following problems: (1) for the experimental evaluation of AG 
practice in PE and sport, at least two experimental and two control groups should be organised 
(low achievers and high achievers in the experimental group, and two heterogeneous subgroups 
in the control group). This definitely enlarges the required number of subjects. (2) The same 
PE teacher (trainer or instructor) has to tutor all the groups in the same environment to ensure 
instruction of an equal quality and type for all the students. This certainly may cause difficulties 
in the experiment. (3) Since the control groups are not classical “passive” groups (they perform 
the curriculum but in heterogeneous groups), significant improvements in the analysed vari-
ables should be expected for both the control and experimental groups. As a result, the authors 
concluded that probable differential effects can only be expected after a substantial period, which 
increases the duration of the experiment. To the best of our knowledge, no study has so far dealt 
with characteristic skills and motor knowledge in SES and studied the effects of AG on them.

54 Ability grouping and improvement in swimming skills Kinesiologia Slovenica, 17, 1, 52–62 (2011) 
Consequently, the aim of the present study was to examine the differential effects of AG versus 
heterogeneous grouping practices on improvement in selected swimming skills. We thus posed 
the following questions: (1) Does AG generally increase the efficacy of the swimming curriculum 
in the improvement of swimming skills? (2) Does AG increase the efficacy of the swimming 
curriculum particularly for the low achievers? (3) Does AG increase the efficacy of the swim-
ming curriculum particularly for the average achievers? (4) Does AG increase the efficacy of the 
swimming curriculum particularly for the high achievers? (5) Do negative consequences arise 
regarding the students’ thoughts (psychological damage) due to the AG?
Apart from the clear lack of studies examining the effects of AG in PE and sport, the authors 
believe that AG, as a possible strategy to improve teaching effectiveness, should be studied in the 
context of swimming particularly due to the well-known high utility of swimming for everyday 
and critical circumstances, which makes swimming one of the most important motor skills and 
motor abilities generally (Brenner, Saluja, & Smith, 2003; Ostrowska, Domaradzki, & Rožek-
Mroz, 2005). 
METHODS
Subjects: Male university students (age: 21 ± 0.9 years; in good health) served as the subject sam-
ple. Their informed consent was obtained before the experiment. The total sample (N = 79) was 
divided into an experimental group (E; n = 37) and a control group (C; n = 42). The groups did not 
differ significantly in the initial (pre-testing) status of their swimming abilities (explained later 
in the text). The C group participated in the swimming curriculum after being randomly divided 
into three heterogeneous subgroups. Prior to this experiment, using discriminant analysis we 
found no significant differences between the C subgroups in their analysed swimming abilities. 
On the contrary, using cluster analysis (K-means clustering) based on the initial swimming 
achievements the E group was clustered in three homogenous subgroups (low, average and high 
achievers), which were significantly different in their analysed swimming abilities. Although 
the results of the discriminant and cluster analyses are not presented in the paper, the authors 
may be contacted for more specific details of these methodological procedures. At the end of 
the programme, based on their initial swimming achievements the C group also was divided 
into three swimming ability-based subgroups (low, average and high achievers). Although the C 
group did not perform the programme in the ability groups, the explained homogenisation in the 
C group was necessary to allow a final comparison of the results achieved by the low achievers, 
average achievers and high achievers from the C and E groups (see the objectives of the study). 
Variables: The sample of variables consisted of 12 swimming skill variables, namely, start, swim-
ming technique, and turn for: (1) freestyle swimming (FS-START, FS-SWIM, FS-TURN); (2) 
breaststroke swimming (BrS-START, BrS-SWIM, BrS-TURN); (3) butterfly stroke (BF-START, 
BF-SWIM, BF-TURN); and (4) backstroke (BaS-START, BaS-SWIM, BaS-TURN). All the vari-
ables were measured by two independent examiners who are swimming teachers using a Likert 
scale ranging from one to five. Following the reliability estimation (see the text below), the 
overall result was calculated as an average of the two observers. All the subjects were tested 
initially (PRE) at the beginning and finally (POST) at the end (in the last week) of the swimming 
programme.
Apart from swimming skills, we observed students’ thoughts regarding the teaching programme 
they had participated in. For this purpose, we used a simplified version of the Questionnaire on 
Students’ Thoughts (Hebert, Landin, & Solmon, 2000). From the originally proposed question-

Ability grouping and improvement in swimming skills 55 Kinesiologia Slovenica, 17, 1, 52–62 (2011)
naire (comprising nine questions), we selected six questions on the basis of the factor validity 
presented in the same paper. Since Hebert et al. used the questionnaire to study tennis, we had 
to adapt it to swimming. After each session, we randomly selected one-third of the participants, 
who answered six questions/statements (presented in Table 5). The scale for each statement was as 
follows: -2 (absolute disagreement); -1 (relative disagreement); 0 (neutral); +1 (relative agreement); 
and +2 (absolute agreement). Namely, the statements were as follows: (1) I was very successful in 
the tasks we practiced today; (2) Today, my performance was good; (3) I’ll never be a good swimmer; 
(4) The tasks we practiced were too hard for me; (5) I really concentrated on the tasks we practiced 
today; (6) I tried very hard to do the best I could. 
Programme: The C group participated in the PE swimming curriculum after having been grouped 
in three heterogeneous subgroups, while the E group performed the PE swimming programme 
in three swimming ability-based subgroups. Generally, the swimming curriculum programmes 
of the E and C groups were equal, meaning: a) each group participated in 60 training sessions 
(45 minutes each; 3 times a week); b) both groups completed the same university swimming PE 
course; c) the objectives of a single training session were the same for both the C and E groups 
(for example, in the two groups the 19
th
 session was aimed at an analytical approach to improve 
the butterfly stroke , the 20
th
 session was aimed at an integrative approach to the butterfly stroke 
etc). Meanwhile, the actual single-lesson routine was adapted according to the needs of the 
individual subgroup. The authors note that the variations of the C and E programmes were not 
planned in advance but were dependent on and induced by the current needs of a particular 
subgroup in each training session. 
Statistical analysis: Apart from the standard descriptive statistics (means and standard devia-
tions), the significance of the differences between the PRE and POST achievements of the C and 
E groups and the three subgroups (low, average and high achievers) were defined using repeated 
measures analysis of the variance (ANOVA). ANOVA was used to determine the differences 
in the variables of students’ thoughts about the programme in which they were participating. 
For the variables of students’ thoughts regarding the teaching programme, when ANOVA was 
found to be significant Tukey’s post-hoc analysis was applied. The level of significance was set 
at 95% (p < 0.05).
R ESULTS
Table 1: Descriptive statistics (Mean and standard deviations – SD) for the PRE and POST meas-
urement in both groups; analysis of the differences within groups (* denotes significant differ-
ences)
TOTAL SAMPLE CONTROL EXPERIMENTAL
PRE POST PRE POST
MEAN± SD MEAN± SD MEAN± SD MEAN± SD
FS-START 3.48±1.14 3.86±0.99* 3.39±1.07 3.67±0.99*
FS-SWIM 3.49±1.13 3.91±1.02* 3.37±1.08 3.64±1.03*
F S -T U R N 3.34±1.23 3.68±1.14* 3.28±1.07 3.64±1.01*
BaS-START 3.30±1.22 3.69±1.00* 3.22±1.04 3.58±0.93*
BaS-SWIM 3.22±1.24 3.63±0.98* 3.20±1.05 3.57±0.97*

56 Ability grouping and improvement in swimming skills Kinesiologia Slovenica, 17, 1, 52–62 (2011) 
TOTAL SAMPLE CONTROL EXPERIMENTAL
PRE POST PRE POST
MEAN± SD MEAN± SD MEAN± SD MEAN± SD
B a S -T U R N 3.16±1.14 3.60±1.02* 3.24±1.02 3.60±0.92*
BrS-START 3.70±1.06 3.97±0.98* 3.33±1.08 3.61±1.03*
BrS-SWIM 3.65±1.08 3.89±1.07* 3.36±1.11 3.62±1.01*
Br S -T U R N 3.65±1.03 3.87±1.05* 3.29±1.08 3.67±0.98*
BF-START 3.36±1.36 3.94±1.01* 3.31±1.09 3.80±1.00*
BF-SWIM 3.12±1.38 3.77±1.06* 3.19±1.13 3.71±0.99*
BF -T U R N 3.30±1.33 4.00±1.02* 3.29±1.09 3.84±0.93*
Legend: Freestyle start - FS-START; Freestyle technique - FS-SWIM; Freestyle turn - FS-TURN; backstroke start - BaS-
START; backstroke technique - BaS-SWIM; backstroke turn - BaS-TURN breaststroke start - BrS-START; breaststroke 
technique - BrS-SWIM; breaststroke turn - BrS-TURN; butterfly stroke start - BF-START; butterfly stroke technique - 
BF-SWIM; butterfly stroke turn - BF-TURN 
Table 2: Descriptive statistics (Mean and standard deviations – SD) for the PRE and POST 
measurement in both groups; analysis of the differences within groups (* denotes significant 
differences) – high achievers
HIGH ACHIEVERS CONTROL EXPERIMENTAL
PRE POST PRE POST
MEAN± SD MEAN± SD MEAN± SD MEAN± SD
FS-START 4.69±0.43 4.81±0.35 4.45±0.68 4.60±0.60*
FS-SWIM 4.81±0.35 5.00±0.00 4.43±0.72 4.58±0.56*
F S -T U R N 4.72±0.42 4.86±0.33 4.28±0.81 4.53±0.66*
BaS-START 4.67±0.56 4.75±0.50 4.17±0.83 4.35±0.68*
BaS-SWIM 4.47±0.64 4.50±0.60 4.13±0.81 4.40±0.75*
B a S -T U R N 4.28±0.51 4.64±0.47* 4.07±0.79 4.27±0.78*
BrS-START 4.81±0.41 4.89±0.33 4.23±0.92 4.42±0.67
BrS-SWIM 4.78±0.42 4.94±0.17 4.32±0.92 4.40±0.65
Br S -T U R N 4.67±0.54 4.81±0.43 4.22±0.93 4.43±0.70
BF-START 4.86±0.33 4.89±0.33 4.40±0.78 4.70±0.48*
BF-SWIM 4.61±0.38 4.83±0.35 4.28±0.87 4.57±0.65*
BF -T U R N 4.78±0.34 5.00±0.00 4.37±0.81 4.65±0.61*
Legend: Freestyle start - FS-START; Freestyle technique - FS-SWIM; Freestyle turn - FS-TURN; backstroke start - BaS-
START; backstroke technique - BaS-SWIM; backstroke turn - BaS-TURN breaststroke start - BrS-START; breaststroke 
technique - BrS-SWIM; breaststroke turn - BrS-TURN; butterfly stroke start - BF-START; butterfly stroke technique - 
BF-SWIM; butterfly stroke turn - BF-TURN 

Ability grouping and improvement in swimming skills 57 Kinesiologia Slovenica, 17, 1, 52–62 (2011)
Table 3: Descriptive statistics (Mean and standard deviations – SD) for the PRE and POST meas-
urement in both groups; analysis of the differences within groups (* denotes significant differ-
ences) – average achievers
AVERAGE ACHIEVERS CONTROL EXPERIMENTAL
PRE POST PRE POST
MEAN± SD MEAN± SD MEAN± SD MEAN± SD
FS-START 3.47±0.99 3.81±1.06* 3.27±0.76 3.55±0.70*
FS-SWIM 3.44±0.97 3.78±0.96* 3.28±0.76 3.60±0.83*
F S -T U R N 3.31±1.01 3.67±1.11* 3.23±0.76 3.63±0.71*
BaS-START 3.08±1.00 3.53±0.87* 3.15±0.65 3.55±0.64*
BaS-SWIM 3.14±0.98 3.53±0.93* 3.10±0.75 3.53±0.65*
B a S -T U R N 3.14±0.98 3.58±0.81* 3.23±0.77 3.60±0.71*
BrS-START 3.86±0.89 4.00±0.82 3.20±0.79 3.55±0.85*
BrS-SWIM 3.92±0.92 4.08±1.03 3.12±0.75 3.48±0.79*
Br S -T U R N 3.78±0.91 3.94±1.00 3.07±0.79 3.53±0.76*
BF-START 3.42±1.14 4.08±0.89* 3.17±0.76 3.65±0.89*
BF-SWIM 3.22±1.3’ 3.97±0.87* 3.02±0.83 3.58±0.79*
BF -T U R N 3.33±1.17 4.17±0.78* 3.15±0.75 3.73±0.72*
Legend: Freestyle start - FS-START; Freestyle technique - FS-SWIM; Freestyle turn - FS-TURN; backstroke start - BaS-
START; backstroke technique - BaS-SWIM; backstroke turn - BaS-TURN breaststroke start - BrS-START; breaststroke 
technique - BrS-SWIM; breaststroke turn - BrS-TURN; butterfly stroke start - BF-START; butterfly stroke technique - 
BF-SWIM; butterfly stroke turn - BF-TURN 
Table 4: Descriptive statistics (Mean and standard deviations – SD) for the PRE and POST 
measurement in both groups; analysis of the differences within groups (* denotes significant 
differences) – low achievers
LOW ACHIEVERS CONTROL EXPERIMENTAL
PRE POST PRE POST
MEAN± SD MEAN± SD MEAN± SD MEAN± SD
FS-START 2.63±0.83 3.25±0.78* 2.41±0.59 2.86±0.81*
FS-SWIM 2.62±0.69 3.27±0.85* 2.36±0.59 2.73±0.77*
F S -T U R N 2.38±0.81 2.87±0.82* 2.29±0.57 2.75±0.81*
BaS-START 2.52±0.93 3.08±0.8* 2.34±0.74 2.88±0.83*
BaS-SWIM 2.42±1.09 3.08±0.87* 2.39±0.75 2.80±0.77*
B a S -T U R N 2.40±0.98 2.98±0.89* 2.41±0.79 2.93±0.82*
BrS-START 2.81±0.69 3.31±0.92* 2.54±0.84 2.84±0.96*
BrS-SWIM 2.73±0.59 3.06±0.78* 2.63±0.96 2.96±1.06*
Br S -T U R N 2.85±0.71 3.15±0.91* 2.55±0.86 3.02±0.99*
BF-START 2.25±0.88 3.17±0.83* 2.36±0.59 3.05±0.83*
BF-SWIM 2.02±0.83 2.91±0.78* 2.25±0.58 2.96±0.81*
BF -T U R N 2.25±0.88 3.21±0.91* 2.36±0.58 3.13±0.82*
Legend: Freestyle start - FS-START; Freestyle technique - FS-SWIM; Freestyle turn - FS-TURN; backstroke start 
- BaS-START; backstroke technique - BaS-SWIM; backstroke turn - BaS-TURN breaststroke start - BrS-START; 
breaststroke technique - BrS-SWIM; breaststroke turn - BrS-TURN; butterfly stroke start - BF-START; butterfly stroke 
technique - BF-SWIM; butterfly stroke turn - BF-TURN 

58 Ability grouping and improvement in swimming skills Kinesiologia Slovenica, 17, 1, 52–62 (2011) 
Table 5. Questionnaire on students’ thoughts regarding success, self-confidence and motivation; 
ANOVA differences
EXPERIMENTAL CONTROL
LA AA HA LA AA HA
SUCCESS
I was very successful in tasks we practiced today 2.67* 3.44 4.01 2.01*
a
3.37 4.56
a
Today my performance was good 2.35 3.27* 4.22 2.23* 3.13 4.39*
SELF-CONFIDENCE (SELF-EFFICACY)
I’ll never be a good swimmer 3.66 3.02 1.99* 4.34*
a
2.76 1.34
a
The tasks we practiced were too hard for me 2.03 2.77 2.01* 3.11
a
2.98 1.55*
a
MOTIVATION
I really concentrated on the tasks we practiced today 3.77 3.43 4.01 2.75*
a
3.7 3.45
a
I tried very hard to do the best I could 3.81 3.90 4.12 3.65 3.45 3.76
a
*denotes a significant difference compared to the other sub-groups in the same programme
a
denotes a significant difference compared to the same ability group in the E programme
LA – low achievers; AA – average achievers; HA – high achievers
In Table 1, results of the overall pre-post differences for the E and C groups are presented. 
Evidently, both groups achieved significant improvements in all of the measured variables. 
Results for the high achievers are presented in Table 2. It is clear that the high achievers of the 
E group improved their swimming skills in nine elements (all of the observed swimming skills 
excluding the breaststroke technique), whereas the high achievers from the C group performed 
significantly better in only one of the observed variables (backstroke turn). 
The average achievers from the C group improved their skills in nine variables from the pre to 
post measurement. At the same time, their peers from the E group performed better in all the 
swimming skills we observed (Table 3). 
There is no evident difference in the improvement of swimming skills for the E and C low 
achievers. Put briefly, both groups improved significantly in all the variables (Table 4). 
DISCUSSION
Ability grouping effects
In the introduction, we specified that one of the possible causes of the deficiency of empirical/
experimental data concerning AG effects in SES can probably be found in the (supposed) fact 
that significant improvements are expectable in the control group. Because both the C and E 
groups improved their swimming skills significantly, this statement is supported by our results. 
Therefore, the possible differential effects of the two programmes cannot be defined accord-
ing to the pre-post differences of the total sample exclusively, but any programme-differences 
should be studied more precisely, most particularly through an analysis of the improvement in 

Ability grouping and improvement in swimming skills 59 Kinesiologia Slovenica, 17, 1, 52–62 (2011)
selected swimming skills separately for the low achievers (LAs), average achievers (AAs) and 
high achievers (HAs). 
One of the main contradictions which follows AG is the so-called “labelling of students,” 
which may communicate self-fulfilling low expectations (Kulik, 1991). Those opposed to AG 
are concerned about the perceived psychological damage to low achievers, the slower pace and 
(potentially) lower quality of instruction, the low expectations for student performance held by 
teachers, and the absence of strong behavioural peer role models in the classes for low-ability 
students. Many “middle level” theorists believe that young adolescents cannot meet the goals 
related to their personal development through AG. They argue that “young people, naturally 
inclined toward learning from their peers, need to be grouped with individuals who are different 
from themselves.” In addition, “adolescents are vulnerable as they struggle to establish a sense of 
their own identity; AG often creates negative perceptions of lower-ability students that affect their 
“self-perceptions”’. Ability grouping, the literature says, has a negative effect on lower achieving 
students’ motivation and opportunities to improve as well as they are able to (adapted after 
Fuligni, Eccles, & Barber, 1995). However, the results of our study do not support the negative 
observations on AG practice, mainly considering the “low achievers”. Quite the opposite, the 
low-achievers evidently benefit from AG in a swimming curriculum as much as their colleagues 
who participated in a heterogenic environment (e.g., both groups improved their swimming skills 
in all measured variables). However (and as previously suggested by Kulik, 1991, and Fuligni, 
Eccles, & Barber, 1995), at the beginning of our experiment the problem of the absence of a 
“model” seemed reasonable so we included one excellent swimmer in each C and E subgroup 
as a model, which probably allowed the low achievers from the E group to avoid any negative 
consequences of the absence of a role model. 
It seems that the high achievers mostly benefitted from the AG. Compared with the initial 
measurement, high achievers from the E group improved nine of their swimming skills, whereas 
their colleagues who participated in the C programme only improved one swimming skill during 
the course. The authors of the paper will briefly discuss possible explanations of such a finding 
from an “insider’s” perspective. Put briefly, when grouped heterogeneously the high achievers 
are rarely in the teacher’s focus, mostly due to the low achievers. A teacher is naturally mostly 
inclined to the needs of those students who are lagging behind and pays special attention to 
their needs. This logically puts high achievers into the background and out of the focus. In our 
study, it was probably even more of an issue because the groups were relatively large (up to 15 
students at the same time). On the contrary, when grouped according to their swimming ability, 
the high achievers were trained according to their specific needs, sometimes even individually, 
allowing them to improve their swimming skills far better in comparison to their peers from 
the C programme. 
The differences in the pre-to-post improvement of the C and E average achievers support all of 
the previously discussed findings. In short, those average achievers who participated in the C 
programme improved all of the measured swimming skills. On the other hand, average achievers 
from the E programme improved all skills excluding the breaststroke variables. Such an incidence 
(e.g., the non-significant improvement of breaststroke skills) probably relates to the fact that the 
breaststroke swimming technique is (as far as the authors of this study are aware) less interesting 
than others (butterfly, crawl and backstroke), mainly because it is the slowest technique of all 
(Zenić & Grčić Zubčević, 2006) and, therefore, the least attractive, while it is also often seen as 
a “female technique.” Both reasons possibly led to students’ decreased interest in improving 

60 Ability grouping and improvement in swimming skills Kinesiologia Slovenica, 17, 1, 52–62 (2011) 
their breaststroke skills. Although we did not study the problem more specifically, previous 
investigations of pupils’ interest and its direct influence on educational achievement support 
our considerations (Koller et al., 2001).
Most of what was discussed above can be supported by the discussion offered by Hebert, Landin, 
and Solmon (2000). Briefly, the authors stated that “when task difficulty exceeds the skill level of 
the learner, practice is typically inappropriate, unsuccessful, and of little apparent value.” Further, 
“individuals who experience failure are likely to develop low levels of self-efficacy, motivation, 
and task persistence, as they do not perceive themselves to be able to master the task.” There 
is no doubt that the low achievers in the heterogeneous groups (the C programme) feel failure 
while competing with their more able colleagues. On the contrary, in the E programme the 
students were initially grouped according to their actual swimming abilities and skills. The 
teacher adapted the training tasks easily according to the students’ (subgroups’) specific needs, 
which led to a superior improvement for the E subgroups. 
Ability grouping consequences
Compared with average- and high-ability students from the same programme, the low-ability 
students from the E and C groups reported significantly lower results for the first question, but 
only those from the C group reported significantly lower levels for the second question regarding 
the personal feeling of success (Table 5). Even more interesting is that the personal feeling of 
success, self-confidence and concentration is significantly higher among the low achievers from 
the E group than those from the C group. Moreover, when observing the differences between the 
low-ability students and their more able colleagues in the C programme it seems that the het-
erogeneous grouping negatively affected the concentration and self-confidence of the low-ability 
students. As a result, we cannot support some previous negative findings concerning AG where 
it was stated that this practice may communicate self-fulfilling low expectations in low-ability 
students (after Kulik, 1991). It is very interesting that we did not establish any significant differ-
ences among the results observed by average achievers from the C and E groups. It seems that 
those students benefit equally from ability and heterogeneous grouping in swimming. Naturally, 
the high-ability students perceived themselves as more successful and self-confident when they 
were grouped with less able colleagues but, according to the previously discussed results, this 
does not necessarily positively affect any improvement in their swimming abilities.
Separately from the previous discussions, certain difficulties arise when teaching students 
grouped according to their ability. They mainly relate to the differentiated instructions which 
should be tailored specifically to each class (group). Although clearly effective, such practices 
put a special emphasis on the teacher (instructor) and demand their constant concentration and 
commitment. However, the results of such approach are evident. 
CONCLUSION
Researchers have struggled for decades to find answers to questions about AG: Does anyone 
benefit from it? Is anyone harmed by it? Who benefits (or is harmed) the most? Why? The answers 
are not always clear-cut and often depend on who you ask and which outcomes are deemed 
important. For many educators, AG is considered a sensible response to academic diversity. For 
others, the practice has harmful unintended consequences and should be abandoned. However, 

Ability grouping and improvement in swimming skills 61 Kinesiologia Slovenica, 17, 1, 52–62 (2011)
according to the presented results grouping on the basis of ability with appropriate differentiated 
instruction is clearly beneficial in the swimming curriculum. For the swimming skills we studied, 
while not statistically observable in general, AG should certainly be considered as a superior 
teaching method for high and low achievers. Most likely, the high achievers benefited from having 
to compete with one another, whereas the low achievers benefited from not having to compete 
with their more able peers. 
Although the results presented in this paper are generalisable on a similar sample of subjects 
(adults and/or adolescent males), there is a certain possibility that we could expect similar effects 
in other samples of subjects. Evidently, the benefits of AG mostly relate to the possibility of 
specifically tailoring an educational programme in which the subjects are involved. However, 
we must not ignore the fact that we studied adult males and it would therefore be hard to expect 
some negative psychological consequences of AG (i.e., they were well aware they should benefit 
from such a practice and that the instructors did not intend to “label” them). Therefore, this issue 
should be noted in future research, especially among younger students.  
REFERENCES
Brenner, R.A., Saluja, G., & Smith, G.S. (2003).Swimming lessons, swimming ability, and the risk of 
drowning. Injury Control and Safety Promotion, 10(4), 211-216.
Bunton, K., Kanihan, S.F., Stacey, F., & Neuzil, M. (2000). Improving media writing with ability groups. 
Journalism and Mass Communication Educator, 55(2), 60-72.
Chambers, R.L. (1988). Legal and practical issues for grouping students in physical education classes. 
Physical Educator, 45(4), 180-186. 
Fredenburg K.B., Lee, A.M., & Solmon, M. (2001). The effects of augmented feedback on students’ percep-
tions and performance. Research quarterly for exercise and sport, 72, 232-242.
Fuligni, A.J., Eccles J.S., & Barber, B.L. (1995). The long term effects of the 7
th
 grade ability grouping in 
mathematics. Journal of Early Adolescence, 15(1), 58 – 89.
Harrison, J.M., Preece, L.A., Blakemore, C.L., Richards, R.P., Wilkinson, C., & Fellingham, G.W. (1999). 
Effects of two instructional models – Skill teaching and mastery learning – On skill development, knowl-
edge, self-efficacy, and game play in volleyball. Journal of Teaching in Physical Education 19, 34-57.
Hebert, E.P., Landin, D., & Solmon, M.A. (2000). The impact of task progressions on students‘ practice 
quality and task-related thoughts. Journal of Teaching in Physical Education, 19, 338-354.
Köller, O., Baumert, J., & Schnabel, K. (2001). Does interest matter? The relationship between academic in-
terest and achievement in mathematics. Journal for Research in Mathematics Education, 32(5), 448-470.
Kulik. J.A. (1991). Findings on ability grouping are often distorted: Response to Allan. Educational Leader-
ship, 48(6), 67.
Linchevski, L.,& Kutscher, B. (1998). Tell me with whom you’re learning, and I’ll tell you how much you’ve 
learned. Journal for Research in Mathematics Education, 29(5), 533-554.
Lou, Y.P., Abrami, P.C., Spence, J.C., Poulsen, C., Chambers, B., & Dapollonia, S. (1996). Within class 
grouping – a meta analysis. Review of Educational Research, 66(4), 423-458.
Ostrowska, B., Domaradzki, J., & Rožek-Mroz, K. (2005) Differentiation and sexual dimorphism of 
children’s morphological features against a background of practising swimming. Kinesiologia Slovenica, 
11(1), 65-78. 

62 Ability grouping and improvement in swimming skills Kinesiologia Slovenica, 17, 1, 52–62 (2011) 
Sariscsany, M.J., & Pettigrew, F. (1997). Effectiveness of interactive video instruction on teacher’s classroom 
management declarative knowledge. Journal of Teaching in Physical Education, 16, 229-240.
Silverman, S., & Skonie, R. (1997). Research on teaching in physical education: An analysis of published 
research. Journal of Teaching in Physical Education, 16, 300-311.
Slavin, R. (1990). Achievement effects of ability grouping in secondary schools: A best-evidence synthesis. 
Review of Educational Research, 60(3), 471-499. 
Slavin, R. (1991). Are cooperative learning and “untracking” harmful to the gifted? Response to Allan. 
Educational Leadership, 48(6), 68-71. 
Škof, B., Cecić Erpič, S., Boben, D., & Zabukovec, V. (2004) Differences in some psychological characteris-
tics between pupils with positive and negative attitudes towards endurance activities in physical education. 
Kinesiologia Slovenica, 10(2), 49-64.
Wang, Y.H., Tung, S.K., Chou, C.C., Huang, M.Y., & Tsai, K.C. (2005). The effects of different tennis teach-
ing models on perceptions of learning environment and learning outcomes of tennis in college physical 
education. Research Quarterly for Exercise and Sport, 76, 92-93 
Wilkinson, S. (1992) A training program for improving undergraduates analytic skill in volleyball. Journal 
of Teaching in Physical Education, 11, 177-194.
Zenić, N., & Grčić Zubčević, N. (2006) Ability grouping – advances in swimming curriculum. Kinesiology, 
37(2), 173-181.