 |
Asia-Pacific Forum
on Science Learning and Teaching, Volume 13, Issue 2, Article 7 (Dec., 2012) |
The aim of study is to investigate opinions of primary students about using performance assessment in science and technology course. This research was carried out in fall semester of 2012. Survey methodology was used in this study. Surveys can be useful when a researcher wants to collect data on phenomena that cannot be directly observed (Karasar, 2005, Cresswel, 2002). Besides, surveys are used to describe attitudes, opinions, behaviours or characteristics of a group (Çepni, 2005).
This study was carried out fall semester of 2012 with 1810 students who educated primary schools in three different provinces of northeast in Turkey. Universe of this study consist of 9820 students and forty primary schools. When sample of study were selected, it was determined that ten schools have so many students in different social economic level. Questionnaire was administered students of these schools educated 4 through 8 grade level and participate in research willingly.
According to gender variable, 53.8% male and 46.2 % female students participated in the study. Grade level variable consist of 27.6 % 4th, 25% 5th, 15.1% 6th, 16.1 7th and 16.1 8th grade. Graduation type of mother variable consists of; 4.5% unschooled, 41% Elementary school, 21% Middle schools, 21% High schools and 12.5% University graduated. According to graduation type of father variable are 2.5% unschooled, 20% Elementary school, 22.5% Middle schools, 32.9% High schools and 22.7% University graduated. In addition to According to monthly family income variable, 35.7 % of students’ families have $250-$500, 33.8 % have $500-$750, 33.8 % have $750-$1000 and 17.7 % have over $1000 income.
In the research, a questionnaire used to collect data consisted of two parts. In the first part, there are some demographic questions as independent variables such as gender, grade level, graduation type of mother and father and family income variable. In the second part, there is a questionnaire which determinate to students’ opinions on performance assessment.
In this study, the questionnaire was developed through the use of five stage model proposed by (Karasar, 2005). In the first stage, many studies related to performance assessment were examined in order to determine the statements about performance assessment and how a questionnaire can be developed (Adanalı, 2008; Algan, 2008; Kanatlı, 2008; Metin, 2008; Metin, 2010; Metin and Demiryürek, 2009; Metin and Özmen, 2009; Metin and Özmen, 2010; Metin, 2011, Metin and Birişçi, 2011; Birgin, 2008, Birgin, 2003; Kan, 2007). After examining, it was carried out interview with 10 students in different grade level and they were asked to five questions about the performance assessment. The five main questions were as follows: 1) “What do you think about effects of performance assessment on students?” 2) “What do students have positive beliefs about performance assessment?” 3) “What do you have negative opinions about performance assessment?” 4) “What do you encounter difficulties to preparing performance task?” and 5) “What do you opinions about portfolio”. These interview and literature helped constitute the item pool.
In the second stage, after interview and reviewing, an item pool was developed which consisted of 50 statements about performance assessment. There were 28 positive and 22 negative statements in the item pool of draft questionnaire. After deciding on the items, an initial item pool was generated and 50 items were put on a five-point rating scale using classifications like “strongly disagree,” “disagree,” “undecided,” “agree” and “strongly agree.”
In the third stage, for the purpose of content validation, an initial draft of the instrument with 50 items on a five-point rating scale was given to a group of four education experts in the fields of language, educational psychology, and educational measurement. Their opinions helped to determine whether the selected items were valid items for assessing students’ opinions about performance assessment. Having received feedback from experts, 15 items were deleted because they were found unsuitable in terms of clarity. According to expert opinions, it was decided that this questionnaire consists of five sub-scales. First sub-scale called on effect of performance assessment on students consists of ten items. Second sub-scale called on positive opinions of students on performance assessment consists of six items. Third sub-scale called on negative opinions of students on performance assessment consists of six items. Fourth sub-scale called on encounter difficulties to preparing performance task consists of five items. Fifth sub-scales called on opinions of students on portfolio consist of eight items.
In the fourth stage, the final draft of the attitude scale with 35 items was administered to 1810 students for calculating validity and reliability of the attitude scale. Students’ responses were entered in an excel file created for further analyses.
In the last stage, the data collected from the 1810 students in the study was analyzed by means of factor analysis and reliability analysis through the use of SPSS 11.5. Before conducting the factor analysis of the scale, the Kaiser–Meyer Olkin (KMO) measurement of sampling adequacy and Barlett’s test were calculated to evaluate whether the sample was large enough to apply a satisfactory factor analysis and was examined to determine appropriateness of factor analysis. The KMO sampling adequacy test statistic was 0.848. This value is higher than the threshold value of 0.01 (Kline, 1994). Barlett’s test of Spherincity statistic was significant [3191.18 (p< 0.01)]. Results of KMO and Barlett’s test appear to support the validity of the factor analysis usage for this study. These five factors of questionnaire explained 65.425% of the total variance. This value is appropriate to consider that other work focused on opinions showed lower explained variance (Kline, 1994: 41%). Besides reliability analysis was performed for each of the emerged sub-scales, and the Cronbach alpha correlation coefficients were used. Then, the Cronbach alpha correlation coefficients were calculated among these factors. It was discovered that general reliable coefficient was 0.92.
Students’ responses to the questionnaire were statistically analyzed according to gender and grade level variables via SPSS 11.5 software. It is just like five point scale and each statement were labeled as 5=completely agree, 4=mostly agree, 3=medium level agree, 2=slightly agree and 1=disagree. Positive attributions were graded as 5-1 and negative attributions were graded as 1-5 questionnaire. Ranges of agreement with the attributions on the questionnaire was determined by using (n-1)/n formula and after calculations the interval width of the range between 1 through 5 was calculated as 0.8. The interval width of 1.00-1.80 showed disagree, the 1.81-2.60 intervals showed slightly agree, the 2.61-3.40 interval showed medium level agree, the 3.41-4.20 interval showed mostly agree, and the 4.21-5.00 interval showed completely agree of agreement with the statements on questionnaire. The mean (x) percentages (%) and frequency (f) scores were computed for each attribution. In the study, some parametric tests such as t-test; one-way analysis of variance (ANOVA) based on p=0.05 significance level were used to clarify the significance of the differences on means. LSD test was used in order to determine source of the differences on means in ANOVA.
