0% found this document useful (0 votes)
19 viewsDocument 9
Research methodology
Uploaded by
Paula BatuoCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
19 viewsDocument 9
Research methodology
Uploaded by
Paula BatuoCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 8
variables to the SPSS Data
Screen
) > OK.
wing sure that dat
ave
or in data entry it ean be checked by using Frequencies or Descriptive
's and necessary correction can be made in data file
From the Freque
box
3. Click oa OK.
Assessing Normality
The assumption of normality canbe verified bythe following graphical techniques:
1. Histogram
Stemand-leaf Plot
es dialogue box, select the variables and move them to Variable(s):
Box Plot
is purpose are
Kolmogoroy-Smimov Statistic witha Lilliefors significance level
The Shapiro-Wilk Statistic
ie and move it to Dependent
ick on Histogram check box snd the Normality plots with tests check box. Select
Factor Levels Together in Boxplots display.
5. Click on Continue and select Both in Explore dialogue box
Use o
8 in Research : An Outline S19
‘6. Click on Options... 1 open Explore: 0
7. Click on Exclude eases pair wise
8. Click on Continue and then OK.
> To Transform Variable (For Normality)
Transform > Compute
In Compute Vari
Options sub dialogue box,
4
lialogue box, select an appropriate variable name.
From the Funetion Group: box select an appropriate transformation (ie. Arithm
1
3
4. From the Functions and Special Vi
5. From the variable lst box, select the variable and press the [HIB button and insert i
the function
6. Click on OK.
» Data Transformation ~ Recoding
1, Transform % Recode > Into Different Variables
2. Inthe Variable List box, select the variable to ecode and cick on [RY 1 move he
‘variable into the Input Variable Output Vari
}. In Name: box type new variable name.
|. Click on Change" Old and New Values...
3.
4
5. Click on second Range: and type a value (the median) in the box.
6. Inthe New Value box, type the new value (ie. 1)
1
8
9.
Clickon Add
Click on third Range: type valuet' (median plus one)
In the New Value box, type the new value (ie. 2).
10. Clickon Add
1. Continue OK
163. DESCRIPTIVE STATISTICS
‘Frequency Distribution - For Continuous Variables (ration / Interval scales) ~ Histogram or
Frequency Polygon
“Measures of Central Tendency and Variability -For Interval or Ratio Data
ies D Frequencies
the variable you require and click on TTR 0 move
at atime)
Analyze Descriptive 8
In Frequencies dialogue box,
the variable into Variable(s): box. (We can select more than one varia 520 Research Methodology with SPSS
7. Clickon Charts... 7
8. In the Frequencies: Charts sub-tialoguc box, selet the charts you require by giving a
click tothe left of names of charts.
9. Click on Continue and thea OK.
From the menu, select File > Save As.
‘= Kar} Pearson’s correlation Coefficient ~ Relationship between two continuous variables
* Phi Coefficient ~Correlation between two dichotomous or categorical variables,
* Poiat-biserial Correlation ~ Correlation between a continuous and a categorical variable
‘© Rank Correlation ~ Correlation between two attributes.
* Related Pairs - Observations must be fom related pairs.
Scale of measurement - Interval or Ratio
ie shouldbe normal
ly related, (Scatter Diagram)
ing of values around the regression line (Seater
Homoscedasticity ~ Uniform
Diagram)
> FOR GETTING A SCATTER DIAGRAM
Click on Graphs menu
ialogue box.
imple Seatter sub-dialogue box,
Select one variable and click on D> button to move the variable into W-axis: box.
Select the second variable and click on the TET] button to move the variable into the
Xeanis: box,
7. Click OK.
1
2
3.
4. Click on Define command pushbutton to open th
5
6
Use of Spss in Research : An Outline $21
y» TO FIND THE KARL-PEARSON’
1, Select Analyze menu
Click on Correlate and then Bivariat vari
oti te. 1 open the Bivariate Correlations dialowue
'S CORRELATION COEFFICIENT
3. Select two variables of requirement and cick on SY
eli to move the variables i
Variables: box, muna
4. Select Pearson from Correlation Coefficients box (at this stage we can
‘Spearman’s Correlation to find Rank correlation coefficient and Kendall's
find the coefficient of association)
From the Test of Significance box, select the One-Tailed ratio batton.
Click Options
In the Statistics pane, select Means and standard deviations by clicking is Check box.
. Click Continue.
Click OK.
> PARTIAL CORRELATION
Suppose we want to find the partial correlation coefficient between the ‘current salary” and
“beginning salary’ of employees at a constant level of ‘Previous experience’ then we have to
follow the following ste
i
find also
-b to
‘lect Analyze > Correlate > Partial.
» in addition to selecting the variables tobe compared; you can also
select Controlling for.
Select Current Salary and Beginning Salary and move them to the Variables pane.
Sctect Previous Experience and move itt the Controlling For pane.
Click Options to select the statistics you want displayed.
Select Means and standard deviations by clicking its check box.
Continue.
Inthe contents pane ofthe output window, click Partial Corr, then clickit again toactivate
it. At the end of the text, type:
Curr Sal X Beg Sal CF Prev Exp.
10. Click anywhere away ftom the ttle to apply the change
‘Note: During your analysis, you're likely to
‘output, Try to come up with a consistent abbreviat
1 glance what you're looking at, In the example above, the X stands f
stands for “Controlling For” $22, Research Methodology with SPSS
10S. T-TESTS
There are three types
1. A One Samp
2. Am Independent Samples test
3. A Paired Samples (Repeated Measures) test
Assumptions
‘Seale of Measurement ~ Interval or Ratio.
Random Sampling - Values selected at random from the population.
+ Normality - Values are normally distributed.
> THE ONE-SAMPLE T-TEST
v
v
1 Select Amalyze menu
2. Click on Compare Means and then One-Sample T Test... to open the One-Sample T
Test dialogue box
3. Select the variable you require ad click on JBI button to move the variable into the
Test Variable(s): box
4. Inthe Test Value: box type the mean score (ji)
5. Click OK.
PAIRED T - TEST (REPEATED MEASURE T - TEST)
1, Analyze Compare Means Paired-Samples T Test.
2. In the Paired-Samples T Test dialogue box, select the variables you require and press
BBP move the variables into Paired Variable: box.
3. Click OK.
INDEPENDENT-SAMPLES T-TEST
+ Before going for indepen:
example. suppose we want
bikes in using ordinary
‘These bikes may ob of
use an independent sampl
‘ordinary petol and speed
tobe included in Factor Li
les t— test, first ensure the data for normality. For
the dependent variables are taken as mileages with
' engine capacity (ecbike) is taken as grouping variable
Before using independent samples t test, we have to ser
= fo sereen the data for normality as
1
2
3.
4
Use of Spss im Research : An Outline $23
Analyze Descriptive Statistics Explore
In Explore dialogue box, sclet the dependent variables and click on Ro move the
variables into the Dependent List: box.
Select the grouping variable and click on to move ito Facto
Click OK.
Conduct the Independent-samples t-test
1
2.
3.
4.
3
6
16.6
>
Analyze Compare Means > Independent SarplesT Text
In the Independent Samples T Test dialogue box, select the test variables and click on
a ee move the variables into the Test Variable(): box
‘Select the grouping variable and click on to move itto Grouping Variable: box.
Click on Define Groups...
In Define Groups sub dialogue box, inthe Group 1: box, type the lowest value of the
‘grouping variable and then press TAB, enter the second value forthe grouping variable in
Group 2: box.
Click on Continue and then OK.
ANALYSIS OF VARIANCE,
ONE-WAY BETWEEN GROUP ANOVA WITH POST-HOC COMPARISONS
Analyze Compare Means" One-Way ANOVA.
the One-Way ANOVA dialogue bor, select the dependent variable and click on
‘Hy co move the variable into the Dependent List: box.
Select the independent variable (Classes) and click on EB to move the variable into the
Factor: box,
‘open the One-Way ANOVA: Options sub dialogue box.
on Descriptive and Homogereity-ofvariance,
Br canteteanc nen Potliae ta ien One Way ANOVA: ot Hoe Mute
ore af eu comeraan pos
able. Select one you want ( preferably Tukey).
Click on Continue and then OK. =
‘TWO-WAY BETWEEN-GROUPS ANOVA ;
1. Analyze General Linear Model > Univariate... open Univariate dislogus box
2. Select the dependent variable and click on [GBI 0 move the variable into the Dependent S24 Research Methodology with SPSS
\ click on JBI t0 move the variable nto the Fixeg
te: Options sub dialogue box.
to open the Univariat
© To plot the cell Means
1. Graphs Line... to open Line Charts dialogue box.
Click on Multiple box and ensure thatthe Summaries for groups of cases is selected
. Click on Define to open Define Multiple Line: Summaries for Groups of Cases sub
disloguebox.
4. In the Line Represent box, select Other Statistic (eg. mean).
5. Selectthe dependent variable and click on RII] to move the variable nto the Variable:
box
6. Select the independent variables with most levels and click on JERI to move the variable
> ONE-WAY REPEATED MEASURES ANOVA
|. Analyze & General Linear Model Repeated Measures... to open the Repeated
Measures Define Factor(s) dialogue box.
Inthe Within-Subject Factor Name: box, type the name for the within subjects factor.
In the Number of Levels: box, type the number of levels of that particular factor.
Click on Add and then Define to open Repeated Measures sub dialogue box.
Select the within subjects variables (e.g. time 1, time 2, time 3, .) and click on By
box.
In the display box, click on the Deseriptive St
Observed power check boxes.
8. Click on Continue and then OK.
> ONE-WAY REPEATED MEASURES ANOVA.
1. Analyze General Linear Model Repeated Measures... to ‘open the Repeated
Measures Define Factor(s) dialogue box.
nL
Use of Spss in Research : An 0: 525
In the Within-Subject Factor Name: box, type the name for the first
factor.
In the Number of Levels: box, type the numberof levels ofthat particular factor.
Clickon Add
In the Within-Subject Factor Name: box, type the name for the second with
factor.
In the Number of Levels: bor type the numberof levels ofthat prt
Click on Add and then Define to open Repeated Measures sub dalo
‘Selectall the combinations of within-subjects variables (
and click on the JGR button to move the variables:
Within Subj bles (A, B): box.
Click on Options... to open Repeated Measures: Options sub dialogue box.
a2b2,...)
within the
). In the display box, click on the Descriptive Statistics, estimates of effect size and
Observed power check bores.
Click on Continue and thea OK.
16.7 FACTOR ANALYSIS
> To Conducta principal axis factor analysis
|. In the Extret box, ensure that
Select the Analyze menu.
Click on Data Reduction and then on Factor. to open the Factor Analysis dialogue
box.
Select the variables you require (i.e. at! 10 att 16) and click on the * Button to move the
variables into the Variables : box.
. Click on the Descriptives... command pusbution to open the Factor Analysis
Deseriptives sub-dialogue box.
In the Statistics box, ensure thatthe Initial solution check box hss beet
In the Correlation Matrix box, select the Coefficients, KMO and B:
‘sphericity and Anti-image check boxes.
Click on continue. ;
Clickon the Extraction...command pushbutton toopen the Factor Analysis : Extraction
sub-dialogue box.
that is displayed inthe box 20.
Research Methodology with SPSS
CClick on the Rotation... command pushbutton to open the Factor Analysis : Rotation
sub-dialogue box.
In the Method box, select the Varimax radio button.
In the Display box, ensure thatthe Rotated solution check box has been selected.
ick on continue.
Click on the Options... command pushbutton to open the Factor Analysis : Options sub.
dialogue box.
In the Coefficient Display format box, select the Sorted by size and Suppress absolute
values less than : check boxes. With the latter check box you are then prompted to enter
the appropriate value you require, whichis usually 3, You will also notice that inthe previous
sub-dialogue box you have the option of replacing missing values withthe mean of each of.
s. This procedure is useful when you have an incomplete data set. By defaul,
missing cases are deleted lstwise, meaning that the whole case is excluded from the analysis.
Click on continue and then OK,
16.8 CLUSTER ANALYSIS
Analyze Classify % Hierarchical Cluster
Select the variables from the list on the left-hand side and transfer themto the box labeled
les by clicking on SOR.
Click on Statistics to open Hierarehi
The main use ofthis dialog box isin specifying a set number of clusters. By default, SPSS
imply merge all cass into a single cluster and it is down tothe researcher to inspect
‘the output to determine substantive sub-clusters. However, if you have a hypothesis about
‘how many clusters should emerge. then you can tell SPSS to create aset number of clusters,
or to create a number of clusters within a range. For this example, leave the default options
"y are and proceed back to the main dialog box by clicking Continue.
ick on Method to open Hierarchical Cluster Analysis: Method sub dialogue box.
Click on an appropriate Method (Choose Ward's method). Inthe below, under Transform
Values select Z - score,
Inthe main dilogue box, select Plots... and then click at Dendrogram
8. Click Continue
1. Once back inthe main dialog box, select
Use of Spss in Research : An Outline $27
ingle solution and then
space, In eality, what we would normally do sto run the ch
this option and them inspect the res
Fun the analysis, requesting that
‘SPSS save coding values for the number of clusters that we identified.
16.9 MULTI DIMENTIONAL SCALING
1
Analyze > Scale > Mi
Multidimensional scaling: Data Format
Select The Data are Proximities.
(PROXSCAL) to open the
Inthe Number of Sourees box, make sure thatthe One Matrix Source is also selected.
1. Inthe One Source box, select The proximities are in a matrix across columns option.
. Click on the Define command pushbutton in open the Multidimensional Scaling
(Proximities in matrices across col... dialogue box.
Select the mine variables representing each ofthe towns and cick onthe JB button to
‘move thevariables into the Proximities : box.
Click onthe Model- command pushbutton o open the Multidimensional Scaling : Model
sub-isloguebox
Inthe Shape : box select theLower-triangular matrix radio buton,
Inthe Proximities : box, select the Dissimilarties radio button.
|. In the Proximity Transformations box, make sure that the Ordinal radio button is
selected,
In the Dimensions : box ensure that you have entered 2 forthe Minimum Dimensions
and 2 forthe Maximum Dimensions
Click on Continue to return tothe Multidimensional Scaling (Proximitis in matrices
across colu.,) dialogue box. - 7
Click on the Plots command pushbutton to open the Multidimensional
sub-dialogue box.
In the Plots: box, ensure that the
distances check boxes are both selected
selected in the Source Plots : Pox
+ Plots
imities vs.
‘Common Space and Transformed Pro
® tected The all soures ali burton should also be S28 Research Methodology with SPSS
Clik on continue. once again to return to the Multidimensional Scaling (Proxiitieg
in matrices across cou.» dialogue box.
16. Click en the Options command pushbutton to open the Multidimensional Scaling ;
Options: sub-
