3. Discriminating Power
The discriminating power of a test item is the ability of test item to discriminate between students with high and low achievement. The
formula used to analyze discriminating power:
D = = P
A
– P
B
Explanation: J = The total number of students who take the test
JA = The total number of students in the upper group JB = The total number of students in the lower group
BA= The total number of in the upper group who got the item right BB = The total number of students in the lower group who got the item right
PA = The porpotion of students in the upper group who got the item right PB = The porpotion of students in the lower group who got the item right
The clasification of the discriminating power D : 0.00
– 0.20 :Poor
D : 0.21 – 0.40
:Satisfactory D : 0.41
– 0.70 :Good
D : 0.71 – 0.100
:Exellent D : negatif
:Discarded
4. Item Difficult
Item diffiuculty is to identify whether a test is categorized into difficult or easy in a test. The formula used to find out the difficulty index as follow:
P =
Explanation: P: Item Difficulty
B: The number of students who got the item right JS: The total number of students who tried the item
Classification of difficulty index: 0.00
– 0.19 = very difficult
0.20 – 0.39
= difficult 0.40
– 0.59 = moderate
0.60 – 0.79
= easy 0.80
– 1.00 = very easy
In analyzing the data, the researcher usedt
test
formula through SPSS Special Package of the Social Sciences version 22 software. The t-
test
is one of a number of hypothesis tests. Before calculated t-test, the researcher did normality and
homogeneity tests first.
5. Normality Test
Before the researcher calculated the value of t-test, the researcher had to analyze the normality and homogeneity of the data. He examinition of
normality was needed to know whether the data has been normally distributed. The researcher used SPSS to tets the normality. In SPSS,
there are two kind of normality test: Kolmogrov Smirnov and Shapiro-Wilk.
The criterion of SPSS: a. If respondents
≥ 50, the normality test uses Kolmograv Smirnov. b. If respondents
≤ 50, the normality uses Shapiro Wilk. The criterion of hypethesis is: