Hey Guys I'm new and I need help

#1
So basically, I have a very little knowledge about using SPSS, but I'm willing to learn.
The problem is I'm having hard time on how to organize the data gathered from the questionnaire to the SPSS.
The questionnaire is based on scaled intensity ( 1- Never, 2 - Sometimes, 3 - Always)
For you to understand easily, Research Title: FACTORS ASSOCIATED WITH THE PERFORMANCE OF THIRD YEAR NURSING STUDENTS IN MEDICAL-SURGICAL NURSING

The questionnaire looks like this:




With this I need to relate the 3 independent variables and the dependent variable
1. Student Related Factors ( Study habits and Sleeping Habits)
2. Learning Environment
3. Teacher Competencies
4. Performance (Dependent) - Grades from Medical Surgical Nursing

Please I need help? I have no money to pay for statistical analysis.

Reps for life.... :D
 

jpkelley

TS Contributor
#2
I would organize this simply in long form:
Code:
       individual test_part component question answer
1  respondent_001        A1     study     A1_1      2
2  respondent_001        A1     study     A1_2      2
3  respondent_001        A1     study     A1_3      1
4  respondent_001        A1     study     A1_4      2
5  respondent_001        A1     study     A1_5      1
6  respondent_001        A1     study     A1_6      1
7  respondent_001        A1     study     A1_7      3
8  respondent_001        A1     study     A1_8      1
9  respondent_001        A1     study     A1_9      3
10 respondent_001        A1     study    A1_10      1
11 respondent_001        A2     sleep     A2_1      1
12 respondent_001        A2     sleep     A2_2      3
13 respondent_001        A2     sleep     A2_3      3
14 respondent_001        A2     sleep     A2_4      2
15 respondent_001        A2     sleep     A2_5      2
16 respondent_001        A2     sleep     A2_6      3
17 respondent_001        A2     sleep     A2_7      1
18 respondent_001        A2     sleep     A2_8      1
19 respondent_001        A2     sleep     A2_9      1
20 respondent_001        A2     sleep    A2_10      3
21 respondent_001         B     learn      B_1      1
22 respondent_001         B     learn      B_2      1
23 respondent_001         B     learn      B_3      1
24 respondent_001         B     learn      B_4      2
25 respondent_001         B     learn      B_5      2
26 respondent_001         B     learn      B_6      3
27 respondent_001         B     learn      B_7      3
28 respondent_001         B     learn      B_8      3
29 respondent_001         B     learn      B_9      1
30 respondent_001         B     learn     B_10      1
31 respondent_001         C     teach      C_1      2
32 respondent_001         C     teach      C_2      2
33 respondent_001         C     teach      C_3      2
34 respondent_001         C     teach      C_4      3
35 respondent_001         C     teach      C_5      2
36 respondent_001         C     teach      C_6      2
37 respondent_001         C     teach      C_7      2
38 respondent_001         C     teach      C_8      1
39 respondent_001         C     teach      C_9      1
40 respondent_001         C     teach     C_10      2
This way, your grouping are hierarchical. Much easier to organize data this way, in my opinion. Is this what you wanted?
 
#4
This way, your grouping are hierarchical. Much easier to organize data this way, in my opinion. Is this what you wanted?
I have 160 respondents, If thats the case 40 X 160= 6400 inputs All in all?
What I want where i can generate Chi-Squares and other statistical data needed for thesis.
Thanks for helping...
 
#5
I'm going to have to disagree with jpkelley and suggest you enter your data in wide format as, with a few exceptions, this format is typically required when analysing data in SPSS. Also, if you're entering the data manually from paper questionnaires, it is slightly less work than long format.

In wide format, every column represents a variable and every row represents a case. Thus your data file will look like this:

Code:
id   a.1.1    a.1.2   a.1.3   a.1.4    a.1.5    ...... c.10
1       1       3        1       2        3     .......   1
2       3       2        1       1        1     .......   3
3       2       2        2       1        1     .......   1
4       2       1        1       3        3     .......   3
5       1       2        2       3        3     .......   2
.
.
160     2       3        2       1        3    .......    1
If you've already begun entering your data in long format, you're going to need to use the restructure wizard to convert it into wide format before you run your analyses.
 
#6
I'm going to have to disagree with jpkelley and suggest you enter your data in wide format as, with a few exceptions, this format is typically required when analysing data in SPSS. Also, if you're entering the data manually from paper questionnaires, it is slightly less work than long format.

In wide format, every column represents a variable and every row represents a case. Thus your data file will look like this:
If you've already begun entering your data in long format, you're going to need to use the restructure wizard to convert it into wide format before you run your analyses.
Youre telling me to consider each question as a variable?
 
#7
For the purposes of inputting your data, yes, each question needs to be treated as its own variable. Once you've entered your data you need to make some decisions about how to go about analysing it. Obviously treating each of the 40 items individually is an unwieldy approach so you'll probably want to look at reducing each of the groups of questions in some way. Without having knowledge of what your exact goals are or the purpose of the analysis it's hard to make a recommendation but a simple approach might just be to aggregate the responses in each group and use these aggregates as your independent variables. As an additional note, you shouldn't combine sleeping and study habits as these are two distinct behaviors.
 

jpkelley

TS Contributor
#8
I'm going to have to disagree with jpkelley and suggest you enter your data in wide format as, with a few exceptions, this format is typically required when analysing data in SPSS.
True, I hadn't seen the reference to SPSS (and the original post wasn't in an SPSS forum). I would therefore agree with helicon on this. That said, the long form reflects the structure typically seen when you combine tables from databases, so I think it's a more intuitive structure for understanding multi-level data. I see your data as multi-level, since you seem to have natural levels of analysis (study habits, etc.). I'm guessing you might also want to divide your questions into groups in order to better interpret your results. Be aware that the wide format is not ubiquitous and may involve restructuring for different type of analyses. Again, for SPSS, go with the advice of helicon.

Double Post
Adrianbrah, can you please include the link for the double post. This will help the moderator fix the issue.
 
Last edited:
#10
Hey, I Guys I have an idea, dunno if this is good.
Given I had 160 Respondents
What if I make a range based on scores. For Study Habits - 30 perfect score.
30 - 20 excellent
19 - 10 average
9 - 1 poor Dunno If the numbers of the range are correct, I need help.(Maybe it needs adjustment)
For example Ive gotten the score as 18, then i belong to average. then i input it to Spss as average, same goes to the other 3 variables
then that would save me some time and effort. Is this possible?
 
#12
Hey, I Guys I have an idea, dunno if this is good.
Given I had 160 Respondents
What if I make a range based on scores. For Study Habits - 30 perfect score.
30 - 20 excellent
19 - 10 average
9 - 1 poor Dunno If the numbers of the range are correct, I need help.(Maybe it needs adjustment)
For example Ive gotten the score as 18, then i belong to average. then i input it to Spss as average, same goes to the other 3 variables
then that would save me some time and effort. Is this possible?
Any advice and comments regarding this idea?
 
#13
Leaving any scale validation issues to the side, the general idea of your approach seems reasonable.

What if I make a range based on scores. For Study Habits - 30 perfect score.
30 - 20 excellent
19 - 10 average
9 - 1 poor Dunno If the numbers of the range are correct, I need help.(Maybe it needs adjustment)
For example Ive gotten the score as 18, then i belong to average. then i input it to Spss as average, same goes to the other 3 variables
then that would save me some time and effort. Is this possible?
However, to be clear, you want to enter all your items and then compute your sum variables within SPSS to maintain the quality of your data - recoding a scale variable into an ordinal variable in this circumstance is the wrong idea. If you code 'Never' as 0, 'Sometimes' as 1, and 'Always' as 2 each participant will have a score ranging between 0-20 - there's no need to further reclassify them into 'excellent', 'average', or 'poor'.
 
#14
However, to be clear, you want to enter all your items and then compute your sum variables within SPSS to maintain the quality of your data - recoding a scale variable into an ordinal variable in this circumstance is the wrong idea. If you code 'Never' as 0, 'Sometimes' as 1, and 'Always' as 2 each participant will have a score ranging between 0-20 - there's no need to further reclassify them into 'excellent', 'average', or 'poor'.
Thanks for taking time to help me, What do you suggest I do?
 
#15
I suggest that you:

1. Enter your survey data into SPSS using the coding recommended above (check that there are no items that need to be reverse coded).
2. Compute subscale totals within SPSS to create your four independent variables.
3. Use multiple regression to model the relationship between your IVs and DV.

Additionally, you haven't mentioned the purpose of your project (homework, work report etc.) so I don't know how rigorous you are expected to be. You may want to explore the relationships amongst items within each of your subscales through reliability analyses and/or exploratory factor analysis.
 
#16
@Helicon
Thanks brah, For time and effort and everything else.

Consulted my teacher and some of my classmates, and learned some of this. (Apparently I wasn't listening)


then transferred to this...

Only 8 respondents were used because this is only the data for the pre-test. I decided to take a shot.
Any tips and remarks?