On-line test is a system of
questions, for which testees give answers by using the Internet and receive results automatically in points simultaneously.
On-line testing is a modern form of control of knowledge and competence in different spheres of
activity. This form of control hasn’t learnt sufficiently yet in terms of theory. Therefore our appeal to
this problem is an urgent challenge.
Advantage of on-line of testing
is that it allows to carry out the fast statistical analysis at a large number of testees, and also replaces a
number of traditional types of level of proficiency control in a foreign
language.
Organization of on-line
testing consists of the following stages:
1. Selection of texts for test
tasks.
2. Task formation.
3. Transfer of the material ACTR/ACCELS
office.
4. Installing test tasks on-line.
5. Distribution of ID to testees and obtaining instructions
on carrying out the on-line
testing.
6. Acquaintance of the testees with the on-line testing instructions.
7. Passing of testing by using PC in a special laboratory.
8. Creation of a matrix in the
SPSS program.
10. Installing test results in the SPSS program.
Selection of texts
plays an important role in the process of creating test tasks since test tasks
for each level depend on texts specifically. That’s why in choosing texts it is
essential to take into account a number of requirements which help to form test
tasks correctly. In creating test tasks for a particular level we shouldn’t
choose a difficult text which, for example, is supposed for higher level.
The second stage is
connected with the process of forming test tasks on the basis of selected
texts, i.e. at this stage test tasks and their distracters are created. To form
tasks we used knowledge gained at seminars and during consultations with qualified
specialists in order to check tasks for accuracy, formulation, cohesion and
other possible technical mistakes.
At the third stage
the created test tasks were installed on-line. Then testees were given out ID
and instructions on testing. At the following stage the testees were acquainted
with the instructions of passing the test.
After the testing
the results were given to the developer. At the following stage all the results
were entered into the SPPS program and statistical data of test tasks were interpreted.
Students of
Foundation Faculty of Al-Farabi Kazakh National University were invited as
testees.
Process of checking
reading skills of foreign students’ who studied at the Foundation Faculty of Al-Farabi
KazNU ran in several stages, first step of which was filling in a special questionnaire for the purpose of getting
from students minimum information that was necessary for receiving more exact
and objective results of testing.
Table 1 – Testees’ age
|
|
Frequency |
Percent |
Valid Percent |
|
|
Valid |
missing |
3 |
12,0 |
12,0 |
|
|
1990 |
6 |
24,0 |
24,0 |
|
|
1991 |
5 |
20,0 |
20,0 |
|
|
1962 |
1 |
4,0 |
4,0 |
|
|
1989 |
4 |
16,0 |
16,0 |
|
|
1986 |
2 |
8,0 |
8,0 |
|
|
1988 |
2 |
8,0 |
8,0 |
|
|
1967 |
1 |
4,0 |
4,0 |
|
|
1992 |
1 |
4,0 |
4,0 |
|
|
Total |
25 |
100,0 |
100,0 |
It is obvious from the Table 1 that age
differentiation of testees lets us to divide them into groups:
First group – testees, who were born in the period of
1962-1967.
Second group – testees, who were born in 1986-1989.
Third group – testees, who were born in 1990-1992.
Table 2 – Gender representation of testees.
|
|
Frequency |
Percent |
Valid Percent |
|
|
Valid |
male |
20 |
80,0 |
80,0 |
|
|
female |
5 |
20,0 |
20,0 |
|
|
Total |
25 |
100,0 |
100,0 |
From the Table 2 it is visible that testees selection
wasn't proportional by gender:
1)
female – 20%,
2)
male – 80%.
In order to know which group – men or women – managed
to avoid more successfully traps of distracters, it was necessary to define
average number of correct answers.
It turned out that share of correct answers per woman
was 7,6 % and per man – 10,1%.
For purity of experiment the students with various languages,
representing different nationalities and being citizens of the different
countries were selected (Afghanistan – 14 people, China – 3 persons, Koreas – 6
people, the USA – 2 persons). Data on the birthplace are presented in the Table
3.
Table 3 – Birthplace
|
|
Frequency |
Percent |
Valid Percent |
|
|
Valid |
China |
3 |
12,0 |
12,0 |
|
|
Afganistan |
14 |
56,0 |
56,0 |
|
|
USA |
2 |
8,0 |
8,0 |
|
|
Korea |
6 |
24,0 |
24,0 |
|
|
Total |
25 |
100,0 |
100,0 |
Six testees’ native language was Dari, for five testees – Korean, for four testees – Farsi, for three –
Pashto, for two – Chinese, for one – English (table 4).
Table 4 – Mother tongues of
testees
|
|
Frequency |
Percent |
Valid Percent |
|
|
Valid |
missing |
4 |
16,0 |
16,0 |
|
|
Persian |
4 |
16,0 |
16,0 |
|
|
Pashto |
3 |
12,0 |
12,0 |
|
|
English |
1 |
4,0 |
4,0 |
|
|
Dari |
6 |
24,0 |
24,0 |
|
|
Korean |
5 |
20,0 |
20,0 |
|
|
Chinese |
2 |
8,0 |
8,0 |
|
|
Total |
25 |
100,0 |
100,0 |
Analysis of gained results showed that higher level of
reading skills in Kazakh belonged to citizens of Afganistan (Table 5).
Table 5 – Correct answers of testees
|
ID |
Number correct |
Number incorrect |
Percent correct |
Country |
|
2679149 |
25 |
15 |
62,50 |
Afganistan |
|
2899457 |
20 |
25 |
50,00 |
Afganistan |
|
3015302 |
19 |
26 |
47,50 |
Korea |
|
8615272 |
17 |
28 |
42,50 |
Afganistan |
|
8976998 |
15 |
25 |
37,50 |
Afganistan |
|
8885146 |
15 |
25 |
37,50 |
Afganistan |
|
6904740 |
15 |
25 |
37,50 |
Afganistan |
|
7823053 |
13 |
27 |
32,50 |
Korea |
|
9583570 |
13 |
27 |
32,50 |
Afganistan |
|
8021950 |
11 |
29 |
27,50 |
Afganistan |
|
3600615 |
11 |
29 |
27,50 |
China |
|
6773400 |
11 |
29 |
27,50 |
Korea |
|
2638173 |
10 |
30 |
25,00 |
USA |
|
6049949 |
9 |
31 |
22,50 |
Afganistan |
|
1719515 |
9 |
31 |
22,50 |
Afganistan |
|
1486085 |
8 |
32 |
20,00 |
Afganistan |
|
1328390 |
7 |
33 |
17,50 |
Afganistan |
|
6529836 |
6 |
34 |
15,00 |
Afganistan |
|
9280675 |
6 |
34 |
15,00 |
Afganistan |
|
1674262 |
4 |
36 |
10,00 |
Korea |
|
3812042 |
4 |
36 |
10,00 |
USA |
|
9956154 |
3 |
37 |
7,50 |
Korea |
|
5288699 |
3 |
37 |
7,50 |
Korea |
|
8610721 |
2 |
38 |
5,00 |
China |
|
8968985 |
2 |
38 |
5,00 |
China |
Successfully
passing the exam doesn’t depend on the length of Kazakh language learning (Table
6).
Table 6 – Duration of Kazakh language
learning
|
|
Frequency |
Percent |
Valid Percent |
|
|
Valid |
missing |
13 |
52,0 |
52,0 |
|
|
1 year |
3 |
12,0 |
12,0 |
|
|
2 years |
1 |
4,0 |
4,0 |
|
|
5 month |
1 |
4,0 |
4,0 |
|
|
6 months |
3 |
12,0 |
12,0 |
|
|
4 years |
2 |
8,0 |
8,0 |
|
|
4 months |
1 |
4,0 |
4,0 |
|
|
Total |
25 |
100,0 |
100,0 |
From the obtained data presented
in table 6 it is visible that testees whose duration of studying of the Kazakh language in 4 years gave 8% of the correct
answers out
of 40, and
the testees who were learning Kazakh language only 6 months managed to answer correctly 12%
of questions from the same quantity of test tasks.
It shows that On-line testing allowed to
optimize both time and a passing form. Besides, on-line testing allowed to
enter quickly all data on results of testing into statistical programs for the
further correlation of statistical
analysis.
1
Balykhina T.M. Osnovy teorii i praktika testirovaniya (v aspekte russkogo
kak inostrannogo): Manual / T.M. Balykhina. – 2 izd., - M.: 2006. – 240.
2
Mayorov A.N. Teoria i
praktika sozdaniya tekstov
dlya sistemy obrazovania. – M, 2000. – 352.
3
Smirnova L.P. Urovni ponimanya inoyazychnykh rechevykh soobshenii na slukh i sposoby kontrolya ikh sformirovannosti (yazykovye i neyazykovye vuzy). Diss. - M., 1982. –
254.
4
Chelyshkova M.B. Teoria i praktika konstruirovaniya pedogogicheskikh
testov: Manual. – M, 2002. – 423p.
5
Lekič M.D. Summit Conference on the Future of Language
Learning. – Washington D.C.: American Council for International Education
ACTR/ACCELS // Russian Language Journal. – Vol. 52-55. – 2004.
6
McNamara T. Language testing. – Oxford: Oxford University Press,
2000. –