An introduction to educational research methods. Введение в образовательные исследовательские методы Білім беру-зерттеу әдістеріне кіріспе

Analysing Quantitative Data

498

age by multiplying by 100 – so 10.4%) that my correlation coefficient of -0.043 would

have arisen by chance when there is no relationship between these two variables.

Although a 10% chance might sound quite low to a newcomer to this approach, in fact

that is still reasonably likely. Consequently, we conclude that there is no evidence to

deviate from the status quo that cognitive development and task motivation are not

related. In fact, a rule of thumb is that in this type of test (a two-tailed test – again,

I won’t say more but you can look this up), you would have to have a probability of less

than 2.5% before you would make a judgement call that the two variables might be

related. Even then we don’t know for sure one way or the other, so you are never able

to couch your conclusions in definitive terms.

2. Tests of difference

As with tests of correlation, there are different types of tests of difference available to

suit the characteristics of the data you have (relating to the measurement level and

distribution of the variable under consideration). If we conduct a parametric test of

difference (because the data are appropriate, as noted in the previous section), then

we compare the mean score of each group on the variable of interest. So, for

instance, if we want to know whether girls and boys differ in their scores on the task

motivation scale, we compare the mean scores of girls and boys on this scale. The

entity calculated in the comparison process is called the t statistic and the test itself

is referred to as a t-test. Here, the assumption underlying the calculations of the

t statistic is that there is no difference between the two groups under consideration

(i.e. boys and girls).

In SPSS, t-tests are accessed via the ‘compare means’ option in the analyse menu.

Although we are conducting a parametric test, there are different ways of comparing

means, and this is reflected in the choice in the ‘compare means’ procedure. Because

we are comparing two different groups (boys and girls) on a particular variable (task

motivation), this is an independent samples test. This is in contrast to a situation

where you might be comparing the same people on two different variables (for instance, in 

my study, I compared students’ scores on each of the motivation scales at the start and end 

of the two-year study to see whether they had changed), which would require the use of a 

repeated measures test. You also need to specify which groups you are comparing (under 

‘define groups’), which you do by inserting the codes you created for the groups you want 

to compare (in my case, 1 and 2, representing boys and girls respectively). The reason you 

need to specify groups is because it may be that you have more than two groups for a 

particular variable in the database (for instance, I might want to compare task motivation 

scores from two of my nine schools). The results of the test are shown in Table 12.6.

Analysing Quantitative Data

499

AnAlysing QuAntitAtive DAtA

203

instance, in my study, I compared students’ scores on each of the motivation scales at 

the start and end of the two-year study to see whether they had changed), which 

would require the use of a repeated measures test. You also need to specify which 

groups you are comparing (under ‘define groups’), which you do by inserting the 

codes you created for the groups you want to compare (in my case, 1 and 2, represent-

ing boys and girls respectively). The reason you need to specify groups is because it 

may be that you have more than two groups for a particular variable in the database 

(for instance, I might want to compare task motivation scores from two of my nine 

schools). The results of the test are shown in Table 12.6.

Table 12.6   Significance test for the difference between boys and girls on the task motivation  

(science) scale

Group Statistics

Gender 

  N 

  Mean 

Std Deviation 

Std Error Mean 

Task 

motivation 

(science) 

Boys 

625 

25.0224 

3.65273 

.14611 

Girls

803

24.9763

3.57090

.12601

Independent Samples Test

Levene’s 

Test for 

Equality of 

Variances

t-test for Equality of Means

95% Confidence 

Interval of the 

Difference

 F.

 Sig

  t

     df

Sig. 

(two-tailed)

Mean 

Difference

Std. Error 

Difference  Lower Upper 

Task 

motivation 

Equal 

variances 

assumed 

.101  .750  .239  1426 

.811 

.04606 

.19240 

-.33136  .42348 

Equal 

variances 

not 

assumed 

.239  1326.515 

.811 

.04606 

.19294 

-.33245  .42457 

The top part of the table shows descriptive statistics, which is handy because we 

hadn’t found out in advance what the mean scores for boys and girls are on the task 

motivation scale (although normally in this sort of analysis you would start by looking 

at the descriptive statistics before conducting statistical tests). As might be expected, 

the mean scores for boys and girls are not exactly the same. In fact, boys (with a mean 

score of 25.02) overall record slightly higher scores than girls (with a mean score of 

24.98), which is not what was expected. To see whether these small differences are 

statistically different, we need to look at the lower part of the table.

13-Wilson-Ch-12.indd   203

8/31/2012   5:41:41 PM

Table 10.6 

Significance test for the difference between boys and girls on the task motiva-

tion (science) scale

The top part of the table shows descriptive statistics, which is handy because we hadn’t 

found out in advance what the mean scores for boys and girls are on the task motivation 

scale (although normally in this sort of analysis you would start by looking at the descrip-

tive statistics before conducting statistical tests). As might be expected, the mean scores for 

boys and girls are not exactly the same. In fact, boys (with a mean score of 25.02) overall 

record slightly higher scores than girls (with a mean score of 24.98), which is not what was 

expected. To see whether these small differences are statistically different, we need to look 

at the lower part of the table.

 

Although there is a lot of information given, for the purpose of this introduction,

you only need to focus on two figures: the t value and the second ‘sig’ (significance)

value. There are two variants for each and the one that is appropriate to interpret

again depends on the characteristics of the data (in this case, whether the amount of

variation in the girls’ scores is as great as that in the boys’, which for these data is a

fair assumption – again, you would need to read up more on this). Hence, assuming

equal variances, we need to focus on the top figure in each case. The calculated

t value is 0.239. Unlike the correlation coefficient, this in itself doesn’t tell you much

about the difference between boys’ and girls’ scores, except to say that the larger this

value, the bigger the difference between the groups. The significance value (0.811),

however, tells us how likely it is that we would have got a t value of 0.239 if there

was no difference between boys and girls. Therefore, in this case, the probability of

Analysing Quantitative Data

500

getting a t value of 0.239 if there was no difference between boys and girls is 81.1%,

i.e. very likely. Using the same benchmark as before (i.e. we would have to have a

probability of less than 2.5% for a two-tailed test), there is clearly no evidence to

depart from the status quo of no difference. So our conclusion would be that there

is insufficient evidence from these data to suggest that boys and girls differ in their

task motivation.

Concluding comments

The purpose of this chapter has been to introduce some statistical techniques,

and to show you how to use a widely available statistical program to do this.

Inevitably, this has been a real rattle through some quite complicated ideas, and

some details have been glossed over. However, hopefully, you feel you’ve grasped

some of the basics and have the confidence to follow these ideas up by consulting

one or more of the texts listed under Further Reading. If so, the chapter has served

its purpose.

Key Ideas

•  Don’t be scared of statistical analysis – even if maths was a mystery at 

school, you can do this – anyone can – you just need to know some basics.

•  Note that use of a specialist program, such as SPSS, means you do not 

need to be a specialist at number crunching, however you still need to 

understand what is going on to ensure you conduct the most appropriate 

analysis to answer your research questions. 

•  Spend time setting up your database properly by allocating an identifier and 

relevant demographic variables and defining characteristics of your variables, 

and allocating values for missing data; and always deal with ambiguous data 

consistently.

•  Always clean your data before starting on statistical analysis to ensure any 

input errors are removed.

•  Start  your  analysis  by  examining  the  frequency  distributions  through 

producing frequency distribution tables and, where appropriate, histograms. 

Look at the relationships between variables measured on scales by plotting 

scatter diagrams.

•  Calculate descriptive statistics to summarize and describe the frequency 

distributions. Only then, use inferential statistics to test whether: there is a 

relationship between two variables, or there is a difference between two 

groups on a particular variable.

Analysing Quantitative Data

501

•  Test relationships between variables using a test of correlation, by calculating 

a correlation coefficient.

•  Test differences between groups on a variable using a test of difference, by 

calculating a t value.

•  Remember that you can only conclude that you have evidence to reject the 

default position (of no relationship/no difference) if your significance figure 

is less than 2.5% for a two-tailed test.

Reflective questions

1.  How will you analyse the data? Think about analysis and plan this before you 

collect the data so you can be sure that the data you gather can answer the 

questions you want to ask.

2.  Which discrete categories (such as low, average and high) will you chose for 

a variable such as ability, from data that has been assessed on a continuous 

scale  (such  as  scores  on  NFER  Cognitive Ability Tests)?  Remember  you 

can’t do the reverse. So, if in doubt, collect data at the highest level of 

measurement possible.

3.  Don’t forget that you tell the computer program what to do. It just follows 

your instructions. So if you put rubbish in, you’ll get rubbish out!

Further Reading

Cohen, L., Manion, L. and Morrison, K. (2011) Research Methods in Education (7th edn).

London: RoutledgeFalmer.

Brace, N., Kemp, R. and Snelgar, R. (2009) SPSS for Psychologists (4th edn). Basingstoke:

Palgrave Macmillan.

Coolican, H. (2009) Research Methods and Statistics in Psychology (5th edn). London:

Hodder & Stoughton.

Field, A. (2009) Discovering Statistics Using SPSS (3rd edn). London: Sage (Introducing

Statistical Methods series). See also http://www.statisticshell.com/html/dsus.html

and accompanying student video clips http://www.sagepub.com/field3e/SPSS

studentmovies.htm (accessed April 2012).

Howitt, D. and Cramer, D. (2007) Introduction to Statistics in Psychology (3rd edn).

Harlow: Pearson Education Limited.

Muijs, D. (2010) Doing Quantitative Research in Education With SPSS. London: Sage.

Sani, F. and Todman, J. (2008) Experimental Design and Statistics for Psychology (2nd edn).

Oxford: Blackwell Publishing.

Analysing Quantitative Data

502

APPENDIX ONE: MOTIVATION QUESTIONNAIRE

AnAlysing QuAntitAtive DAtA

207

Appendix one: motivation questionnaire

Name: ....Ros McLellan……………… 

Date of Birth:      …25/07/96....................… 

School: …KTS……………………… 

Science Teacher:  …class 1………..................……

Date: …...13/09/07………………… 

Boy or Girl:         …girl……………….......................

I am going to ask you some questions about your experience of school. There are no right or wrong answers. Please 

answer all the questions as honestly as you can. Your teachers will not see your answers.

For each question decide whether you:  1 – strongly disagree,  2 – disagree,    3 – neither agree or disagree,  4 – agree, 

or  5 – strongly agree.

For each question cross the answer that is closest to what you think.

For example: 

Baked beans taste good  

1     X     3     4     5

This person doesn’t really like the taste of baked beans. That is why 2 is crossed, which is disagree.

Another example: 

Tottenham Hotspur are great   

1     2     3     4     X

This person is a big Spurs fan. 5 is crossed, which is strongly agree.

First of all I want you to think about WHEN you feel you’ve had a really SUCCESSFUL day at school.

Please answer the following questions and remember:

1 – strongly disagree,  2 – disagree,  3 – neither agree or disagree, 

4 – agree,  5 – strongly agree.

1)  I feel successful when I don’t have to work hard

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

2

X

4

5

2)  I feel successful when I learn something interesting

In school generally

In science

In English

1

2

3

4

X

1

2

3

X

5

1

2

X

4

5

3)  I feel successful when I show people that I’m clever

In school generally

In science

In English

1

X

3

4

5

1

2

X

4

5

1

2

X

4

5

13-Wilson-Ch-12.indd   207

8/31/2012   5:41:43 PM

Analysing Quantitative Data

503

SCHOOL-BASED RESEARCH

208

Remember: 

1 – strongly disagree,  2 – disagree, 

3 – neither agree or disagree,  4 – agree,  5 – strongly agree.

4)  I feel successful when I do almost no work and get away with it

In school generally

In science

In English

X

2

3

4

5

X

2

3

4

5

X

2

3

4

5

5)  I feel successful when people don’t think I’m thick

In school generally

In science

In English

1

2

3

X

5

1

2

3

X

5

1

2

3

X

5

6)  I feel successful when I solve a tricky problem by working hard

In school generally

In science

In English

1

2

3

X

5

1

2

3

X

5

1

2

3

X

5

7)  I feel successful when I’m the only one who can answer the teacher’s questions

In school generally

In science

In English

1

2

3

X

5

1

2

3

X

5

1

2

3

4

X

8)  I feel successful when I don’t have to do any homework

In school generally

In science

In English

1

X

3

4

5

1

2

X

4

5

1

2

X

4

5

9)  I feel successful when I do well without trying

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

X

3

4

5

10)  I feel successful when a lesson makes me think about things

In school generally

In science

In English

1

2

3

X

5

1

2

3

X

5

1

2

3

X

5

13-Wilson-Ch-12.indd   208

8/31/2012   5:41:43 PM

Analysing Quantitative Data

504

AnAlysing QuAntitAtive DAtA

209

Remember: 

1 – strongly disagree,  2 – disagree, 

3 – neither agree or disagree,  4 – agree,  5 – strongly agree.

11)  I feel successful when I tell a teacher a fib and get away with it

In school generally

In science

In English

X

2

3

4

5

X

2

3

4

5

X

2

3

4

5

12)  I feel successful when I don’t have any difficult tests

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

X

3

4

5

13)  I feel successful when I do better work than other pupils

In school generally

In science

In English

1

2

X

4

5

1

2

X

4

5

1

2

X

4

5

14)  I feel successful when I don’t do anything stupid in class

In school generally

In science

In English

1

2

X

4

5

1

2

X

4

5

1

X

3

4

5

15)  I feel successful when I work hard all day

In school generally

In science

In English

1

2

3

4

X

1

2

3

4

X

1

2

3

4

X

16)  I feel successful when I mess around and get away with it

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

2

3

X

5

17)  I feel successful when I get higher marks than other pupils

In school generally

In science

In English

1

2

3

4

X

1

2

3

4

X

1

2

3

4

X

13-Wilson-Ch-12.indd   209

8/31/2012   5:41:43 PM

Analysing Quantitative Data

505

SCHOOL-BASED RESEARCH

210

Remember: 

1 – strongly disagree,  2 – disagree, 

3 – neither agree or disagree,  4 – agree,  5 – strongly agree.

18)  I feel successful when teachers don’t ask me any hard questions

In school generally

In science

In English

1

2

X

4

5

1

2

X

4

5

1

2

X

4

5

19)  I feel successful when I get good marks on a test without studying

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

X

3

4

5

20)  I feel successful when I finally understand a really complicated idea

In school generally

In science

In English

1

2

3

4

X

1

2

3

4

X

1

2

3

4

X

21)  I feel successful when other pupils get things wrong and I don’t

In school generally

In science

In English

1

X

3

4

5

1

X

3

4

5

1

X

3

4

5

13-Wilson-Ch-12.indd   210

8/31/2012   5:41:43 PM

Analysing Quantitative Data

506

Analysing Quantitative Data

507

АНАЛИЗ КОЛИЧЕСТВЕННЫХ ДАННЫХ

жүктеу/скачать 32,4 Mb.

Достарыңызбен бөлісу: