Analysing language test data

#1
Could somebody help me with interpreting the results of the following survey?

I have 25 main respondents in the survey. They are non-native speakers of English and they have been doing a grammar test in English with 50 cloze questions. Some days later, the respondents were presented the same exercises as multiple choice questions, these including all the responses they had given on the first version with the addition of the correct answer, if this hadn't been amongst their responses. In addition, thereafter 25 native speakers of English did the first version of the test.

I will present the data below.

1st version / 2nd version (multiple choice) for non-native speakers
33 / 27
31 31
32 29
30 33
29 31
29 30
28 30
29 23
27 30
28 29
27 26
26 27
25 24
25 20
24 25
24 24
23 32
23 23
22 24
21 15
22 24
21 23
17 18
16 17
16 24

1st version for native speakers
36
35
33
34
34
33
33
32
32
32
32
31
31
31
31
31
31
30
30
29
28
28
26
25
24

What do you think, are these two sets of scores somehow related (non-native speakers)? How do I find out the correlation coefficient and how can I interpret it (non-native speakers)? What is the relationship between non-native speakers' scores and native speakers' scores on the first version of the test?

How could I analyse this data in statistical analysis programme R?

Thank you for your answers!
 

JohnM

TS Contributor
#2
The Pearson (r) correlation coefficient can be computed between the 1st and 2nd versions. Here is some reference material:

http://faculty.vassar.edu/lowry/ch3pt1.html

You can compare the native speakers' scores both sets of scores for non-native speakers either by a 1-way ANOVA or separate t-tests (info on these can be found on this site).

I don't have any experience with the R language.