2007; 29: e198–e203
WEB PAPER
PBL curriculum improves medical students’ participation in small-group tutorials Y. T. WUN, EILEEN Y. Y. TSE, T. P. LAM & CINDY L. K. LAM Family Medicine Unit, The University of Hong Kong, Hong Kong
Abstract Background: Group learning is the core of problem-based learning (PBL) but has not been extensively studied, especially in
Asian students. Methods: This study compared students of PBL and non-PBL curricula in students’ talking time and participation in small-group tutorials in a medical school in Asia. The proportions of student talking of 46 tutorials in three teaching rotations of the PBL curriculum and those of 43 corresponding tutorials in the non-PBL curriculum were counted. Twelve videotapes of tutorials (six from each curriculum), stratified for tutor, case scenario and students’ learning stage, were randomly selected and transcribed. They were rated with the group-interaction (5 items) and active-participation (four items) tutorial assessment scales developed by Valle et al. These outcomes were compared between the students of PBL and non-PBL curricula. Results: Students from the PBL curriculum talked significantly more. In only two (4.7%) of 43 tutorials in the non-PBL curriculum did the students talk more than the tutors; but students talked more than the tutors in 17 (37.0%) of 46 tutorials in the PBL curriculum curri culum.. PBL students scored significantl significantly y highe higherr than non-PBL non-PBL stude students nts in all items excep exceptt one item (resp (respect ect to peers peers)) of the tutorial assessment scales, and in the mean scores of both the group interaction scale (items 1–5) and the active participation scale (items 6–9).
Conclusions: The results suggested that PBL starting from the early years of a medical curriculum was associated with more active
student participation, interaction and collaboration in small-group tutorials.
Introduction
Practice points
Problem-b Proble m-base ased d lea learni rning ng (PB (PBL) L) ena enable bless stu studen dents ts to lea learn rn group-work skills and attitudes, and improves their communication skills (Wood 2003). These skills and attitudes include teamwork, cooperation, respect for colleagues’ views, chairing a group, and interaction with group members (Wood 2003). Hence PBL is thought to be good for group learning (Dolmans et al. 2005). PBL medical graduates are more likely than nonPBL graduates graduates to indi indicate cate that they have learn learned ed commu communicanications skills and teamwork (Prince et al. 2005). However, there is not much research evidence on the effect of PBL on group learning, especially among students in Asia. In fact, one study observed that undergraduates in psychiatry had no difference in their learning style whether they were in a PBL or non-PBL curriculum (McParland et al. 2004). It is often difficult to get students to work in a group or as a team (Distlehorst (Distlehorst et al. 2005; Walton et al. 1997). There is also a general impression that group learning is even more difficult in As Asia ian n st stud uden ents ts wh who o ten tend d to be pa pass ssiv ive e an and d un unwi will llin ing g to cha challe llenge nge oth others ers.. Our pre previo vious us stu study dy on con consec secuti utive ve videotaped tutorials of medical students in Hong Kong in 1995 (Dixon et al. 1997) seemed to support this hypothesis. We found that students were not much engaged in discussion, which was connected mainly through the tutor. In one third of the tutorials, the students spoke less than 30% of the time,
Asian Asian stu studen dents ts fro from m a non non-PB -PBLL med medica icall cur curric riculu ulum m participated little in discussion in small-group tutorials. PBL in intr trod oduc uced ed in th the e ea earl rly y ye year arss of a me medi dica call . PBL curriculum curri culum could enabl enable e Asian students to parti participate cipate more actively in small-group tutorials. . Objective criterion-based assessment of student performance in small-group tutorials is feasible. . Further research on the effect of a PBL curriculum on the quality of student discussion and learning is required. .
more than 40%. In only two out of 58 tutorials did the students spoke spo ke mor more e tha than n the tuto tutors. rs. Thi Thiss pat patter tern n was irr irresp especti ective ve of wh wheth ether er the tut tutori orials als wer were e con conduc ducted ted in the stu studen dents’ ts’ mother tongue (Cantonese) or in English. On the other hand, Kember & Gow (1996) argued that the apparent stereotype of passi passive, ve, super superficia ficiall and rote learn learning ing behaviour behaviour of Asian studen stu dents ts mig might ht be exp explai lained ned by the tea teachi ching ng met method hod and environment rather than inherent characteristics. Since September 1997, our medical school adopted PBL starting from year one of the curriculum. This provided an opportunity to compare students of the new PBL-curriculum with those of the traditional non-PBL curriculum in their performance in small-group tutorials. The aim of the study was
in ano anothe therr one third 31– 31–40% 40%,, and the rem remain aining ing one thi third rd Correspondence: Cindy L. K. Lam, Family Medicine Unit, the University of Hong Kong, 3rd Floor, Ap Lei Chau Clinic, 161 Main Street, Ap Lei Chau,
Hong Kong. Tel: 852-25185653; fax: 852-28147475; email:
[email protected]
e198
ISSN 0142–159X print/ISSN 1466–187X online/07/060198–6 2007 Informa UK Ltd. DOI: 10.1080/01421590701294349
PBL improves student participation
evalua luate te the eff effect ectss of PBL on stu studen dentt par partic ticipa ipatio tion n and to eva interaction in small-group tutorials.
Methods Study setting and subjects Before 1997, our medical school had a traditional curriculum thatt was lec tha lectur ture e and dis discip ciplin line e bas based. ed. The cur curric riculu ulum m was revised with PBL playing a significant part since the first year in 1997 but the student selection process had not changed. Since 1994, our final-year medical students during their family medicine clerkship had weekly small-group (seven to eight students) tutorials based on scenarios of common problems in family practice. These problem-based tutorials differed from the usual PBL tutorials in that the problems were mostly dealt with in one session (for clinical analysis and management) with the main aim for the students to learn problem-solving skil sk ills ls an and d a se seco cond ndar ary y ai aim m of ge gene nera ratin ting g ne new w le lear arni ning ng objecti obj ectives ves.. The fam family ily med medici icine ne cle clerks rkship hip con contin tinued ued the proble pro blem-b m-base ased d tut tutori orials als wit with h the sam same e pro proble blem m sce scenar narios ios until 2002 when the first batch of students of the new PBL curriculum had the family medicine clerkship.
Data collection Forty-eight tutorials over three clerkship rotations (with four student groups in each rotation and four tutorials per group) in 2002 (PBL group) were videotaped. Two trained research assi as sist stan ants ts,, af after ter st stan anda dard rdiz izat atio ion n on th thre ree e ta tape pes, s, vi view ewed ed sepa se para rate tely ly th thes ese e ta tape pess an and d re reco cord rded ed wh wheth ether er it wa wass th the e student or tutor talking at each one-minute interval. The oneminu mi nute te in inter terva vall wa wass us used ed af afte terr a pi pilo lott tes testt sh show owin ing g th that at counting at half-minute intervals was technically difficult and gave similar results to those obtained obtained at one-m one-minute inute intervals. intervals. The tutorials lasted for different lengths of time and there were silenc sil ence e at som some e one one-mi -minut nute e sto stops. ps. The stu studen dentt and tutor talkin tal king g pro propor portio tions ns of eac each h tuto tutoria riall wer were e exp expres ressed sed as the percentages of the total talking incidents. Silent incidents were nott co no coun unted ted.. Th The e me meth thod od of co coun unti ting ng th the e pr prop opor orti tion on of
of 24 items in four scales: independent study, group interaction, reasoning skills, and active participation. The four scales had demonstrated internal construct validity with scale inter-item correlation corre lationss rang ranging ing from 0.59–0 0.59–0.88, .88, and adequ adequate ate inter internal nal reliab rel iabili ility ty Cro Cronba nbach ch alp alphas has of gre greate aterr tha than n 0.8 0.8.. (Va (Valle lle et al. 199 1999) 9) Since Sin ce we aim aimed ed to ass assess ess onl only y stu studen dents’ ts’ par partic ticipa ipatio tion n and coll co llab abor orat atio ion, n, we us used ed on only ly th the e ra ratin ting g sc scal ales es on gr grou oup p intera int eractio ction n and act active ive par partic ticipa ipatio tion, n, whi which ch con consis sistt of nin nine e items as shown in Appendix A. Items 1 to 5 formed the group interaction scale (ability to communicate and fit into the group) while items 6 to 9 formed the active participation scale. Each item was rated on a 6-point Likert scale from ‘1’ (never) to ‘6’ (always). A pilot study was done to confirm the applicability of Valle’s items and to standardize the assessment criteria on the transc tra nscrip ripts ts and vid videot eotape apess of two tut tutori orials als tha thatt wer were e not included in the main study were each assessed by the rater of the main study and a teacher who was experienced in PBL teaching. teachi ng. Valle’s items were found to be applicable applicable and there was good inter-rater correlation in the scores. The 12 tutorials included in the main study were all rated by a trainer in Family Medicine who had extensive experience in underg und ergrad raduat uate e tea teachi ching ng but had not taught taught in any of the curric cur ricula ula und under er stu study. dy. He fir first st rat rated ed the per perfor forman mance ce of the students on all the items except item 4 by reading through through the transcripts transcripts in which the identity of the year (1995 or 2002) was concealed. Since item 4 (listens attentively to other members of the group) could not be rated from the transcripts, the rater reviewed the videotape to score the students on this item after he had scored all the other items. The scores of the other items were not to be altered.
Data analysis Students’ participation in the tutorials was measured by the studen stu dents’ ts’ tal talkin king g pro propor portio tions. ns. We exp expecte ected d tha thatt stu studen dents’ ts’ talking proportion would be less than 40% in two-thirds of the non-PBL groups (Dixon et al. 1997). Hence, we compared the proportion of tutorials between the two study years by student talking proportions of <20%, 21–30%, 31–40%, 41–50% and above 50% to see if the students from the PBL curriculum
studen stud entt or tu tuto torr ta talk lkin ing g wa wass th the e sa same me as th that at us used ed in ou our r previous previ ous study in 1995 (Dixon et al. 1997). Similar Similar data on the propor pro portio tions ns of stu studen dentt and tutor tal talkin king g in the tut tutori orials als of three matched clerkship rotations in the academic year of 1995 (non-PBL group) were retrieved from the data collected in our previous study (Dixon et al. 1997). Students’ group interaction and participation were assessed qualita qua litativ tively ely in sixtutori sixtutorialsrando alsrandomly mly sel select ected ed fro from m eac each h of 199 1995 5 and 2002 batches stratified for the tutor, case scenario, and studen stu dentt sta stage ge of tra traini ining ng (i. (i.e. e. the stu studen dents ts hadunder hadundergon gone e sim simila ilar r specialty clerkships before the tutorials took place, to minimize the effects of knowledge on students’ performance in group tutorials). The videotapes of these tutorials were transcribed in verbatim and each student’s performance was assessed with a rating form adapted from Valle’s tutorial assessment question-
made any improvement. We used the Fisher’s Exact test to analyse the statistical difference between the two groups. ForValle’ss ass ForValle’ assess essmen mentt sca scales les,, we ana analys lysed ed the sco scores res of eac each h item and scale as ordinal data and used the Mann–Whitney U test to detect any statistical difference between the non-PBL and PBL stu studen dents. ts. We als also o com compar pared ed the mea mean n sco scores res by tut tutori orials als,, in which we took the mean of the scores of all students in each tutoria tuto riall as the tut tutori orial al sco score. re. Man Mann–W n–Whit hitney ney U testwas als also o use used d to analyse the difference between tutorials from the two years. Two-tailed p -values -values <0.05 were considered statistically significant. We used the Cronbach alpha to evaluate the internal consistency of the items in each scale. All data analyses were done by the SPSS for Windows 11.0 programme (SPSS Inc. Chicago).
naire and validated byof a team of physicians trained in PBLdeveloped teaching methods as part a project to evaluate PBL outcomes at the National Autonomous University of Mexico (Valle et al. 1999). The original Valle’s assessment form consists
Results Out of 48 tutorials in Year 1995, the data on the proportion of stude stu dent nt ta talk lkin ing g co coul uld d be re retr trie ieve ved d in on only ly 43 43.. Of th the e 48
e199
Y. T. Wun et al.
videotaped tutorials of Year 2002, one tutorial was not completely compl etely recorded and the sound quality quality in another was too poor for assessment, leaving 46 tapes in the final analysis. Forty-f For ty-four our students students in 199 1995 5 and 42 stu studen dents ts in 200 2002 2 wer were e assessed on 12 videotapes (six randomly selected from each year) with Valle’s scales.
(a) Talking proportion in a tutorial Table 1 shows the distribution of the proportions of student talking by years. In only two (4.7%) tutorials did the students of 1995 talked more than the tutors; but students of 2002 talked more mo re th than an th the e tu tuto tors rs in 17 (3 (37. 7.0% 0%)) of th the e 46 tu tuto tori rial als. s. The difference was statistically significant (P < 0.001, Fisher’s Fisher’s Exact Test). Students talked more than 40% in 16 (37.2%) out of the 43 tutorials in 1995, while students talked more than 40% in 29 (63.0%) out of 46 tutorials in 2002, the difference was statistically significant (P ¼ 0.01, Fisher’s Exact Test).
or one-tailed) for item 3 (adjusts to different group roles), item 6 (helps his/her peers to clarify ideas), and item 7 (participates in case discussions). The tutorials of 2002 had higher mean scores for both the group interaction scale (items 1–5) as well as the active participation scale (items 6–9) than the tutorials of 1995. The differences were statistically significant. The mean scores of students of 2002 were higher those of students of 1995 in all items except item 5 (respect to peers) and bot both h sca scales les,, wh when en the they y wer were e ana analys lysed ed as ind indivi ividua duals. ls. In contrast to the analysis by tutorials (Table 2), the differences between betwe en students of the two study years reached statistical statistical significance (Table 3) in all items except item 5. The mean scores for item 5 were similar for students of both years. The Th e in inte tern rnal al co con nsi sist sten enci cies es of th the e Va Vall lle’ e’ss gr grou oup p intera int eractio ction n (ite (items ms 1–5 1–5)) sca scale, le, the act active ive par partici ticipat pation ion sca scale le (items 6–9), and all 9 items as a whole were assessed. For all 86 students, the Cronbach alphas for group interaction, active part pa rtic icip ipat atio ion, n, an and d al alll 9 it item emss we were re 0. 0.40 40,, 0. 0.70 70,, an and d 0. 0.72 72,, respectively.
(b) Interaction and participation in tutorials Table 2 shows the tutorial assessment scores of the tutorials by clerkship clerk ship year. The tutor tutorials ials of 2002 scored higher than those of 19 1995 95 in al alll it item emss ex exce cept pt it item em 5 (r (res espe pect ct to pe peer ers) s).. The differences were statistically significant (either two-tailed
Discussions This study evalu evaluated ated the perfo performanc rmance e of students in small small-group tutorials objectively and cross-validated the results with
Table 1. Distribution of talking proportions in tutorials. Number (%) of 1995 tutorials ( n ¼ 43) Talking proportions
Teacher talking
Number (%) of 2002 tutorials ( n ¼ 46)
Student talking
Teacher talking
Student talking
0–20%
0
4 (9.3%)
0
2 (4.4%)
21–30%
0
8 (18.6%)
1 (2.2%)
4 (8.7%)
31–40%
0
15 (34.9%)
4 (8.7%)
11 (23.9%)
41–50%
2 (4.7%)
14 (32.6%)
13 (28.3%)
12 (26.1%)
51–60%
16 (37.2%)
2 (4.7%)
14 (30.4%)
13 (28.3%))
61–70%
14 (32.6%)
0
11 (23.9%)
3 (6.5%)
71–80%
9 (20.9%)
0
1 (2.2%)
1 (2.2%)
81–90%
2 (4.7%)
0
2 (4.3%)
0
>90%
0
0
0
0
Table 2. Tutorial mean Valle’s tutorial assessment item and scale scores.
Year
#1
#2
#3
#4
#5
#6
#7
#8
#9
Group interaction Active participation scale (#1–#5) Scale (#6–#9)
1995 ( n ¼ 6)
Mean (SD) Ran ge ge
1.33 (0.817) 1 .0 .0– 3.0 3.0
1.68 (1.182) 1.0– 1.0 – 4.0 4.0
2.01 (0.818) 1.0 –3 1. –3.0
4.27 (0.452) 3.6 –4 3. –4.8
3.68 (0.312) 3.3 .3– – 4.0 4.0
2.44 (0.312) 1.3– 1.3 –3 .4 .4
4.10 (0.211) 3 .9 .9– 4.4 4.4
2.44 (0.752) 1.6– 1.6 –3.4
2.21 (0.950) 1.5– 1.5 –3.9
12.98 (2.020) 1 1.0 1.0– –16 .3 .3
11.19 (2.047) 8.7 .75 5–1 3.8 3.86 6
2002 ( n ¼ 6)
Mean (SD) Rang Ra nge e
2.14 (0.994) 1.0– 1. 0–3. 3.9 9
3.14 (1.565) 1.0–5.0 1.0–5 .0
3.10 (0.401) 2.4–3. 2.4– 3.6 6
4.76 (0.492) 4.4–5.7 4.4–5 .7
3.57 (0.823) 2.1– 2. 1–4. 4.4 4
3.57 (0.823) 2.9–4.3 2.9–4 .3
4.79 (0.235) 4.4–5. 4.4– 5.1 1
3.24 (0.942) 1.4– 1. 4–4. 4.1 1
3.00 (0.571) 2.3– 2. 3–4. 4.0 0
16.71 (1.740) 14.4 14 .43– 3–19 19.7 .71 1
14.71 (1.454) 12.7 12 .71– 1–17 17.0 .00 0
MWU
7.5
8.0
4.0
8.0
3.0
0.5
8.0
7.5
2.0
2.0
P
0.073
0.102 0.
0.024 0.
0.107 0.
0.016 0.
0.005 0.
0.107 0.
0.092 0.
0.010
0.010
15.5 0.686
SD ¼ standard deviation; MWU ¼ Mann–Whitney U statistic; P ¼ P value for Mann–Whitney U test.
e200
PBL improves student participation
Table 3. Student mean Valle’s tutorial assessment item and scale scores.
#1
#2
#3
#4
#5
#6
#7
#8
#9
Group interaction scale (#1–#5)
Active participation scale (#6–#9)
Sum Mean (SD)
60 1.36 (0.810)
73 1.66 (1.219)
87 1.98 (1.562)
189 4.30 (0.851)
162 3.68 (0.639)
106 2.41 (1.530)
180 4.09 (1.074)
108 2.45 (1.532)
96 2.18 (1.603)
571 12.98 (2.538)
490 11.14 (4.067)
Range
1–4
1–4
1–5
2–6
2–5
1–6
2–6
1–5
1–5
9–19
5–21
2002
Sum
150
205
217
389
312
261
381
244
222
1273
1108
( n ¼ 42)
Me an an (SD) Range
2.14 (1.389) 1–4
3.14 (1.523) 1–6
3.10 (1.605) 1–6
4.76 (0.932) 2–6
3.57 (1.085) 2–6
3.69 (1.297) 1–5
4.79 (1.094) 3–6
3.24 (1.411) 1–5
3.00 (1.593) 1–5
16.71 (3.330) 11–25
14.7 (3.744) 9–21
MWU
654.0
461.5 46
608.5 60
667.0 66
887.5 88
471.5 47
6 04.0 60
668.0 66
675.0 67
346.0
484.5
P
0.004
<0.001
0.003
0.018 0.
0.723 0.
<0.001
0.004
0.019 0.
0.021 0.
<0.001
<0.001
Year 1995 ( n ¼ 44)
SD ¼ standard deviation; MWU ¼ Mann–Whitney U statistic; P ¼ P value for Mann–Whitney U test.
both quantitative and qualitative methods. The same rater was used in the qualitative assessment of student performance in all tutorials tutorials in order to achie achieve ve greater consistency consistency in scoring. The Th e us use e of tw two o ra rate ters rs fo forr ea each ch ta tape pe mig might ht im impr prov ove e th the e reliab rel iabili ility ty of the results results but it was very dif diffic ficult ult to ide identi ntify fy anothe ano therr per person son who was experien experienced ced in PBL and had not
for further exploration with students. One possible explanation could cou ld be tha thatt our students students wer were e rel reluct uctant ant to cha challe llenge nge or confront each other because it is considered socially improper to do so in the Asian culture. In Val Valle’ le’ss ori origin ginal al stu study dy all the ass assess essmen mentt sca scales les had Cronbach alphas above the group comparison standard of 0.7.
participa partic ipated ted in the two cur curric ricula ula.. Oth Other er stu studie diess on rat rating ing of videotaped consultations showed that a single rater could achiev ach ieve e an acc accept eptabl able e rel reliab iabili ility ty (Ra (Ram m et al. 199 1999), 9), wit with h gene ge nera rali liza zabi bili lity ty co coef effi fici cien entt of 0. 0.82 82 fo forr ra rati ting ng 12 ta tape pess (Hays (Ha ys et al. 2002). Although Although there cou could ld be per person sonal al bia bias, s, we tried to minimize this by blinding and the use of a structured assessment form. The bias, if any, would be more likely to affect the absolute scores and should be similar for both groups of students. We aimed to find out the relative difference between PBL and non-PBL students rather than the absolute scores of individual students, therefore the bias of the rater should not have affec affected ted the concl conclusion usion of the results. In this study, students of the PBL curriculum, compared with students of the non-PBL, had more active participation and interaction among themselves during small-group tutorials as sho shown wn by the obs observ ervati ation on tha thatt the they y tal talked ked mor more e dur during ing
The alpha of the group interaction scale (items 1 to 5) was 0.83 butt th bu this is sc scal ale e sh show owed ed su subb-op opti tima mall in inter terna nall re reli liab abil ilit ity y (Cronbach alpha ¼ 0.40) in our study, which suggested that the items might not measure the same domain. We carried out factor analysis on the scores of these five items and found three components: item 1, items 2 þ 3, and items 4 þ 5, but the Cronbach alphas were still low (0.50 and 0.53 respectively for the last two components). We could not find in MEDLINE and EMBASE EMBAS E other studies on the psychometrics psychometrics of Valle’ Valle’ss tutori tutorial al assessment asses sment scales except the original; original; theref therefore ore we cann cannot ot be certain whether this finding was related to the instrument or unique to our students. Further research is needed to confirm the validity and reliability of Valle’s tutorial assessment scale on group interaction. Until more information on the psychometrics of this scale is available, the results from these five items ite ms sh shou ould ld be in inte terp rpre rete ted d in indi divi vidu dual ally ly in inste stead ad of as a
tutorials. This was further supported by the evaluation with Valle’s assessment scales, particularly in the aspects of adjustment of individual roles in the group, helping peers to clarify ideas/opinions, and active participation in discussion. We tried to evaluate the group function during tutorials and foun fo und d th that at tu tuto tori rial alss of PB PBLL st stud uden ents ts ha had d hi high gher er me mean an assessment scores than those of non-PBL students (Table 2). Beca Be caus use e of th the e sm smal alll sa samp mple le si size ze of 12 tu tuto tori rial als, s, th the e improvement in the tutorial assessment scores did not reach statistical significance except in three items. When the students were assessed as individuals, the improvement reached stat st atis isti tica call si sign gnif ific ican ance ce du due e to a la larg rger er sa samp mple le si size ze of 86 (Table 3). The majority of the mean scores for the 1995 nonPBL students were below 3.0 (lower half of the scale) out of the Likert scales of 1–6. On the contrary, the mean scores for
summative scale. The results results of our study sup suppor ported ted Kember and Gow Gow’s ’s hypothesis that teaching method could mould Asian students’ learning behaviour (Kember & Gow 1996). The change from a traditional to a PBL curriculum in our medical school seemed to hav have e cha change nged d the beh behavi aviour our and att attitu itudes des of Chi Chines nese e students in tutorials and enabled them to be more active and collaborative. Although it was possible that the difference in students’ performance between the tutorials in 2002 and those in 1995 coul co uld d be th the e re resu sult lt of a di diff ffer eren ence ce in th the e tr trai aini ning ng an and d experience of the tutors in the two years independent of any effect from the curricula, this was unlikely because the same thre th ree e ac acad adem emic ic sta staff ff we were re th the e tu tuto tors rs of mo more re th than an th thre ree e quarters of the tutorials in both years and all of them had
th the e sidere 2002 20 02d PB PBL stud st uden ents ts . we were respect abov ab 3.0 3. 0rs’th that at co coul uld donl be consid con ered asL sat satisf isfact actory ory. ‘Respe ‘Re ctove toe pee peers’ was the only y item in which students did not show any improvement. We had not looked into the cause of this and this could be a topic
com comple pleted ted formal for mal PBL and sma small-g ll-grou roup learni lea rning ng tra traini ining ng courses at McMaster University before thep1995 tutorials took place. Furthermore our earlier study (Dixon et al. 1997) did not find any difference between tutorials by the tutor who had
e201
Y. T. Wun et al.
more than 15 years of experience and those by the tutors who had 3 to 5 years of experience. This study suggested that PBL starting from the early years of the curriculum could improve the group learning process. PBL stude students nts demon demonstrate strated d better collaboration collaboration in discu discussion ssion by adjus adjusting ting their indiv individual idual roles duri during ng grou group p discu discussion ssion,, help he lpin ing g th thei eirr co coll llea eagu gues es to cl clar arif ify y id idea eas/ s/op opin inio ions ns,, an and d participating more in discussion, as indicated by significantly higher high er scores in items 3, 6, and 7 of Valle’s assessment assessment scales. scales. Though Thou gh th this is st stud udy y di did d no nott me meas asur ure e th the e ef effe fect ctiv iven enes esss of le lear arni ning ng di dire rect ctly ly,, ot othe hers rs ha have ve sh show own n th that at le lear arni ning ng is more mo re ef effe fect ctiv ive e if th ther ere e is co coll llab abor orat atio ion n am amon ong g gr grou oup p memb me mber erss (P (Pai aice ce & He Hear ard d 20 2003 03). ). Th Ther eref efor ore e we be beli liev eve e that th at st stud uden ents ts of th the e PB PBLL cu curr rric icul ulum um ar are e mo more re li likel kely y to have ha ve mo more re ef effe fecti ctive ve le lear arni ning ng th than an stu stude dent ntss of no nonn-PB PBLL curriculum.
Limitations This is a cross-sectional study with a historical-control group instead of a randomised controlled trial, which might not have contro con trolle lled d for all the con confou foundi nding ng fac factor tors. s. Alt Althou hough gh bot both h groups were final-year students with similar clinical exposure, students of seven years apart might have acquired different learning learn ing beha behaviour viourss due to diffe difference rencess in paren parental tal attitu attitudes des and changes in the secondary school education approach. approach. There could also be cultural changes in both society and the medical school. This study focused on student participation in small-group learning and did not explore some other important components nen ts of PBL suc such h as ide identi ntifyi fying ng kno knowle wledge dge gap gaps, s, set settin ting g learni lea rning ng obj objecti ectives ves,, ind indepe epende ndent nt inf inform ormati ation on sea search rch,, and sharing new information. The study did not assess the nature and quality of students’ talking, an increase in the quantity of partic par ticipa ipatio tion n in sma smallll-gro group up tut tutori orial al did not nec necess essari arily ly impl im ply y th that at th the e qu qual alit ity y of th the e co cont nten entt of di disc scus ussi sion on,, wa wass also better.
Conclusion This study compared the performance in small-group tutorials betwee bet ween n stu studen dents ts of a PBL curricul curriculum um and students students of a traditiona tradi tionall nonnon-PBL PBL curriculum curriculum in a medica medicall schoo schooll in Asia. We found that PBL students talked significantly more in the tutori tut orials als tha than n non PBL stu studen dents. ts. Qua Qualit litati ative ve ass assess essmen ment t also als o sho showed wed tha thatt PBL stu studen dents ts wer were e mor more e int intera eractiv ctive e and partic par ticipa ipator tory y in tuto tutoria rials. ls. PBL sta starti rting ng fro from m the ear early ly yea years rs of a me medi dica call cu curr rric icul ulum um mi migh ghtt en enab able le As Asia ian n me medi dica call
e202
students to be more collaborative and active in small-group learning. Further research is required to confirm the results of thi thiss his histor torica icall cas case e con contro troll stu study dy and to det determ ermine ine the effect eff ect of PBL on the quality quality of dis discus cussio sion n in sma small ll gro group up tutorials.
Notes on contributors All authors are Specialists in Family Medicine and members of the Family Medicine Unit, The University of Hong Kong. Y. T. WUN was responsible for the rating of tutorials, and a main contributor in data analysis and the writing of this paper. E. TSE contributed in the planning and supervision of the project, data analysis, and writing of this paper. T. P. LAM contributed in the planning of the project and writing of this paper. C. L. K. LAM contributed to the planning of the project, data analysis, and writing of this paper.
References Distlehorst LH, Dawson E, Robbs RS, Barrows HS. 2005. Problem-based learning outcomes: the glass half-full. Acad Med 80:294–299. Dixon AS, Lam CLK, Lam TP, Ho R. 1977. Can Hong Kong students cope with small-group, problem-based learning? in: J. Conway, R. Fisher, L. Sher Sheridanidan-Burn Burnss & G. Ryan Ryan.. (Eds (Eds), ), Research Research and Devel Developme opment nt in Problem-based Learning , Vol. 4, pp. 119–126, (Newcastle, Australia, The Australian Problem-based Learning Network). Network) . Dolmans DHJM, De Grave W, Wolfhagen IHAP, Van Der Vleuten CPM. 2005. 200 5. Prob Problem-b lem-based ased learn learning: ing: futu future re chall challenge engess for educ education ational al practice and research. Med Educ 39:732–741. Hayss R, Spike N, Gup Hay Gupta ta TS, Hollis Hollis J, Vei Veitch tch J. 200 2002. 2. A per perfor forman mance ce assessment module for experienced general practitioners. Med Educ 36:258–260. Kember D, Gow L 1996. A challenge to the anecdotal stereotype of the Asian student. Stud St ud High Educ 161:117–12 161:117–128. 8. Mcparland M, Lorraine NM, Gill L. 2004. The effectiveness of problembased base d learn learning ing compa compared red to tradit traditional ional teaching in unde undergrad rgraduate uate psychiatry. Med Educ 38:859–867. Paice E., Heard S. 2003. Collaborative learning. Med Educ 37:758–759. Prince Pri nce KJA, Van Eij Eijss PWL PWLJ, J, Bos Boshui huizen zen HPA, Van Der Vleuten Vleuten CPM CPM., ., Scherpbie Sche rpbierr AJJA. 2005 2005.. Gene General ral comp competenc etencies ies of prob problem-b lem-based ased learning (PBL) and non-PBL graduates. Med Educ 39:394–401. Ram P, Grol R, Rethans JJ, Schouten B, Van Der Vleuten C, Kester A. 1999. Assessment of general practitioners by video observation of communicative and medical performance in daily practice: issues of validity, reliability and feasibility. Med Educ 33:447–454. Valle R, Petra I, Martinez-Gonzalex A, Jose Antonio Rojas-Ramirez Ja, Morales-Lopez S, Pina-Garza B. 1999. Assessment of student performance man ce in pro proble blem-b m-base ased d lea learni rning ng tut tutori orial al ses sessio sions. ns. Med Ed Educ uc 33:818–822. Walton JN, Clark DC, Glick N. 1997. An outcomes assessment of a hy hybr brid id-P -PBL BL co cour urse se in tre treat atme ment nt pl plan anni ning ng.. J De Dent ntal al Ed Educ uc 61:361–367. Wood DF. 2003. Problem based learning. Brit Med J 326:328–33 326:328–330. 0.
PBL improves student participation
Appendix x A: Vall Valle’s e’s tutorial Appendi assessment rating scales on group interaction and active participation (Some items were further defined for standardization in this study, as shown in the brackets) Group interaction scale 1. 2. 3.
4. 5.
Accepts suggesti Accepts suggestions ons abou aboutt his/her his/her work Acce Ac cept ptss de deci cisi sion onss ma made de by th the e gr grou oup p (‘ (‘Ye Yes’ s’,, no counter-decisions, excluding counter-suggestions) Adjusts Adjus ts to different different group group roles roles (Leader (Leader role: role: changes changes discussion discu ssion topic; helps to defin define e learn learning ing objective. Facilitator role: e.g. ‘So, . . .’ ) List Li sten enss at atten tentiv tivel ely y to ot othe herr me memb mber erss of th the e gr grou oup p (Eye contact, body language of attentiveness) Shows Sho ws respect respect to his his/he /herr pee peers rs (Acknowl (Acknowledg edge e oth other er students’ stude nts’ contr contributio ibution, n, no interr interruption uption,, no perso personal nal attack)
Active participation scale 6.
Helps his/he his/herr peers to clarify clarify ideas ideas (Immedia (Immediate te followfollowup statem statements ents to clarif clarify y anoth another er stude student’s nt’s statement without tutor’s facilitation. Gives further factual knowledge.) 7. Par Partici ticipat pates es in cas case e dis discus cussio sions ns (In (Initia itiates tes que questi stions ons// directions/ direct ions/topic topicss of discu discussion ssion,, exclu excluding ding respo response nse to tu tuto tor’ r’ss di dire rect ct qu ques estio tions ns.. Jo Join inss in di disc scus ussi sion on,,
8.
9.
e.g.,, e.g. Tuto Tu torr-St Stud uden ent1t1-St Stud uden ent2 t2-S -Stu tude dent nt77-Tu Tuto tor. r. Suggestive Sugge stive utterance, e.g., ‘yeh’ ‘yeh’,, ‘ah’, ‘yes’. Exclu Exclude de behaviours described by Item 8 and Item 9) Shares knowledge with the group (Gives his/ hi s/he herr pr prev evio ious us re read adin ings gs.. Vo Volu lunt ntee eers rs hi his/ s/he her r knowledge) Givess fee Give feedba dback ck (Si (Silen lences ces a dis disord ordere ered d dis discus cussio sion, n, e.g., e.g ., ‘sh ‘shih’ ih’.. Rem Remind indss oth others ers abo about ut the dis discus cussio sion n direction)
e203