Computer analysis of research data is of no value without suitable programme. A programme is a set of instructions which a computer employs to analyze data. There are large number of programmes available for data analysis, each appropriate for different statistical computations.
The researcher needs to anticipate the method of data analysis and data output needed in his research work. This may involve only one statistical programme or several programmes before the desired analysis is completed. The researcher can obtain info about existing programmes from computer centers.
For analysis of data in social science to which educational research belongs is the statistical package for the social sciences(SPSS).SPSS contains statistical procedures like mean, standard deviation, correlation coefficient, t-test, chi square, regression analysis, ANOVA, ANCOVA, and other advance statistics.
Others software includes; UCLA Biomedical Series(BIMED) for applications in the biological and medical field, SAS(statistical analysis system), Minitab, R, e.t.c.
SPSS stands for Statistical Package for the Social Sciences and is a software package used to manage and analyse quantitative data.One of the advantages of SPSS is that students can import data from other sources, when data is organized as a database, including Excel. Importing an Excel spreadsheet to SPSS for the data analysis is a fairly simple process, requiring some preparation and a few basic steps.As a matter of fact,you can simply copy and paste data directly from Ms. Word into SPSS.
SPSS allows researchers to analyze data without using advanced statistical analysis skills and it has a simple user friendly interface.You don’t need to be a guru in Mathematics or advanced Statistics to use to analyze your quantitative data.
Data collected from questionnaire, achievement test or other instruments in quantitative research methods have to be analyzed and interpreted using statistical procedures found in SPSS.
When you successfully start the SPSS program from the start menu,you will notice that the SPSS data file(.sav) has two views. Namely, the Data View and Variable View as shown below:
1. The data view
This is where you enter responses. Note that each column represents a variable(var) , each row represents a case and cells contain values. You can click on each image to show a larger, detailed version. Click anywhere on the image to close it and return to this page.
2. The variable view
This is where you will define your variables — essentially the questions that you have asked your participants in your questionnaire, interview etc. The first row is used to create an ID number for each individual participant (omit real names for anonymity and ethical reasons). Each row thereafter is one of your questions, so you can see here that row 2 is a question about the gender of the participant, row 3 is about the age of the participant, row 4 about their academic qualification,etc.
Referring to the image above, you need to define information about each variable, the main ones you will need to define are summarized in the following table:
|Variable.||Some notes||For example for “gender”|
|Name||– Up to 8 characters, no spaces |
– Could relate to the question number (e.g. Q1) or be descriptive (e.g. Gender)
|Type||e.g. numeric, date, string (alphanumeric or letters)||Numeric|
|Width||The number of characters that can be entered||2|
|Label||Description of the variable (i.e. a longer description of the variable name)||Gender of respondent|
|Values||– Labels that explain the values – Numeric values are preferable |
– Be consistent e.g. always 1 = no, 2 = yes
|1 = male 2 = female|
|Columns||Defines the width of the column in data view||7|
|Align||Left, right or centre alignment of data in cells||Right|
|Measure||Nominal, ordinal or scale (interval/ratio)||Nominal|
CHECKING, CODING AND SETTING-UP(Rostering) DATA FOR SPSS AND DATA ENTRY
Questionnaire checking–The first step to preparing data for analysis is to check the questionnaires if they are accepted or not. You will check if the filling of the questionnaire is complete or not. This is done during the data collection period or later on.
The checked questionnaires will be rejected and remove on the basis of the following points:
- Unqualified participant filling the questionnaires
- If the instruments are fully or partially not answered
- If it is answered in such a way that shows that the respondent could not answer the questions.
There will be need for you to collect more data when too many questionnaires are rejected.
Questionnaire coding-this is the process of assigning numerical value to the variables. As a researcher you are at liberty of devising coding procedure but it should be one appropriate to the data collected.
One can code subject names, sex, socio economic status, qualification, occupation, income e.t.c. These are mainly categorical variables which can be replaced by numbers.
Examples of data coding are given below:
- Sex 1-Male
- Socio Economic Status 1-High SES
- Programme 1-NCE
Note that consecutive numbers are used in the coding above. The last number in the series is used for “other” or miscellaneous when necessary
This is the procedure of recording the coded data for use in a computer analysis.
For instance: Suppose a researcher used thirty students in a study which involves checking the effect of three teaching methods on students’ learning.
The variables involved are as follows:
- 10 students who were taught with lecture method only
- 10 students who were taught with discussion method only
- 10 students who were taught with both lecture and discussion methods
Other moderator variables are gender, achievement level and socio-economic status.
To roster or set such data for analysis, the researcher will first of all, code it in this manner.
|Subject No.||Treatment||Sex||IQ||Socio economic status (SES)|
|01-Musa 02-Eze 03-Gbenga 04-Moses||1-Lecture Method 2-Lecture & Discussion method 3- Discussion Method||1-Male |
|1-High 2-Low||1-High SES 2-Average SES 3-Low SES|
The first items in the roster above are designated by code while the remaining two are actual data collected.The rostered data above is good for key punching and computer analysis.
In subsequent posts,we shall learn how to analyze our data after preparing them as discussed in this post.