Scribd
Upload
6 views9 pages

SPSS Statistical Analysis Techniques

This document provides an overview of various statistical tests and procedures in SPSS, including reverse coding, t-tests, ANOVA, regression, correlation, and cross tabulation. Reverse coding is used to address response bias by reversing the scoring of some survey items. T-tests are for comparing two groups on a continuous variable, while ANOVA is for comparing three or more groups. Regression examines the relationship between continuous variables. Correlation measures the association between two variables. Cross tabs generate a 2x2 contingency table to analyze relationships between two categorical variables.

Uploaded by

Vividh Bansal
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
6 views9 pages

SPSS Statistical Analysis Techniques

This document provides an overview of various statistical tests and procedures in SPSS, including reverse coding, t-tests, ANOVA, regression, correlation, and cross tabulation. Reverse coding is used to address response bias by reversing the scoring of some survey items. T-tests are for comparing two groups on a continuous variable, while ANOVA is for comparing three or more groups. Regression examines the relationship between continuous variables. Correlation measures the association between two variables. Cross tabs generate a 2x2 contingency table to analyze relationships between two categorical variables.

Uploaded by

Vividh Bansal
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

SPSS Helpdesk

Reverse Coding

Is used to remove acquiesce Bias. Reverse coding is achieved in an inventory when the normal flow of the spirit of the item is
reversed.

For eg:

In an inventory comprising of 5 items if 4 items are extolling the virtues of the respondent and one item is criticizing the item is
reverse coded.

t-TEST: Conducted for difference between 2 groups. For 3 or more we use annova

eg: M Vs F: t-Test

If 1st Sig > 0.05 we can


assume equal variances.
(We reject Null hypothesis).
Levene’s Test: No difference
b/w variance:: We assume
equal variance

There is a difference in
data sets: (1-0.514 =
0.49)::49%<95% which
implies null hypothesis
is void. Therefore we
can say that there is a
difference b/w m-f as
far as empowerment is
concerned.
 Select Step variable
 Select Grouping variable
 Go for second table in op sheet
 Check for sig of levenes test : IF sig>0.05 ::Accept first row ::equal variances assumed

Else Equal variances not assumed

 In either case go for sign-2 table 4 whichever row selected if value > 0.05 then there is a difference in perception about
the test variable by the grouping variable.

For t-Test with: Test Variable : Continuous Variable

Grouping Variable : Categorical variable

Note: Catagorical variable has no decimal points.


Annova:

If we try and compare b-w 3 or more categorical variable based on their perception then (Continuous variables) then we go for
Annova (one way annova)

1. Go to analyse
2. Mean-> one way annova
3. Dependant list : Continuous Variable Factor : LOM (level of management)
4. Options -> Discriptive

(1-0.126 =87%) => as it is <95%


therefore is significant difference.
Regression: When we go for multivariate stats we must go for continuous variables

Correlation

Regression

Simple Linear Multiple Linear Structural Modeling

Used for Causal problems.

Eg. Organizational Culture-> is caused by unified values, beliefs etc…

1. Analyze
2. Regression
3. Linear regression
4. Dependant :
5. Independent :

Data used here is incorrect…Kindly get the data from Ruchira

1. Model Summary: Adjusted R Square : Is the amount of


variation caused in the dependent var coz of
independent variables. (Explained -> Unexplained)
involvement can explain 31% of variation in
performance.
2. Coefficients: sig has to be <= 0.05 if so then causal
relation between the two
Bivariate Analysis :

1. Analyze
2. Correlate -> Bivariate
Here we use Organizational Culture is dependent on HR Practices

One Tailed Test/Two Tailed Test:

One Tailed : Only on achievement. When one aspect can tell abt other

Two TAILED: Where both aspects are known


Step1: Check Sig Value

Step 2: See Adjusted R


Square
Mutivariable Regression

Stepwise: Will intro


variable one at a time

Forward: Start from


Least to most influential
Variable
K-S Test: Test weather data is normal or not.

 Analyze
 Non-parametric Test
 K-S Test

KS Data is normal

Cross Tabs: 2X2 data eg. Johari Window

 Analyze
 Discriptor
 Cross Tabs

You might also like