Tutorial IBM SPSS Statistics
TIXXXX Statistika Multivariat
Ni Luh Putu Lilis Sinta Setiawati
23418015
Materials
1 Statistical Test 2 Data Treatment
• Normality test • Outliers data
• Homogeneity test
3 Discriminant 4 Correspondence
Data Types
1 Nominal 2 Ordinal
(male and female) (Likert scale)
3 Interval 4 Ratio
(temperature) (weight, height)
Normality Test
1 Shapiro-Wilk 2 Kolmogorov-Smirnov
(parametric test) (non-parametric test)
Decision criteria
Significance < 0.05 normal distribution
Significance > 0.05 non-normal distribution
Kolmogorov-Smirnov
Analyze – Non-parametric test – Legacy dialogs – 1-Sample K-S
Normality Test
Decision
X1 non-normal;
X2, X3, Y normal.
Homogeneity Test
Decision criteria
Significance < 0.05 non-homogenous
Significance > 0.05 homogenous
*only required for some parametric test
Homogeneity Test – Levene Test
Analyze – Compare means – One way ANOVA – Options – Homogeneity of variance
Homogeneity Test
Decision
Y data set is homogenous
Outliers Data
Could disturb data distribution to non-normal distribution
Analyze – Descriptive statistics – Descriptive – Variables
– Save standardized value as variables
SPSS Output
Outliers Data
Definition (Sufren & Natanael, 2013)
Outliers data define as Z-score comes outside the range of -2.5 and +2.5
If there exist, you should delete the data.
Discriminant Analysis
Definition
• Determine the probability of group membership
• Independent variables are predictors; dependent variables are groups
Assumptions
• No outliers
• Predictors should be normally distributed
• Equality variance across groups
• Multicollinearity (predictor variables cannot be highly correlated)
(find more in the internet)
Discriminant analysis
Analyze – Classify – Discriminant – Grouping variables – Independents –
Define range (0 and 1)
Discriminant Analysis
Discriminant Analysis – Output Interpretation
Discriminant Analysis – Output Interpretation
Discriminant Analysis – Output Interpretation
Correspondence Analysis
Function
To examine the relationship between two nominal variables graphically
in multidimensional space.
Example: examine the relationship between product and fragrances. If
the product is shampoo, what kind of fragrance is one’s expecting?
Fragrances
Product
Rose Jasmine Sandal Lilly Menthol Lavender Total
Bath soap 148 40 314 30 106 52 690
Shampoo 42 186 10 86 232 42 598
Body lotion 170 116 58 40 36 88 508
Hair oil 20 226 26 50 126 38 486
Deodorant 94 64 98 40 50 154 500
Face cream 200 46 116 48 24 56 490
Total 674 678 622 294 574 430 3272
Correspondence Analysis
1) Data – Weight cases – Weight cases – Frequency variable (score)
2) Analyze – Dimension Reduction – Correspondence analysis – Row
(product) – Column (fragrances) – Define range (1 – 6) – Update
– Plots (biplot, row points – column points)
Correspondence Analysis – Output Interpretation
Correspondence Analysis – Output Interpretation
Correspondence Analysis – Output Interpretation
Correspondence Analysis – Output Interpretation
Correspondence Analysis – Output Interpretation
Related links
• Normality: https://www.youtube.com/watch?v=42K7w1Fs4Dk
• Homogeneity: https://www.spssindonesia.com/2014/02/uji-homogenitas-dengan-spss.html
• Outliers: https://www.youtube.com/watch?v=6MmZ4ZSIpDs
• Discriminant analysis:
https://www.youtube.com/watch?v=vm1OLoL_VuY
https://www.statistikian.com/2013/12/interprestasi-analisis-diskriminan-dengan-spss-part-2.html
https://www.statistikian.com/2013/12/analisis-diskriminan.html
https://www.statistikian.com/2013/12/interprestasi-analisis-diskriminan-dengan-spss.html
• Correspondence analysis: https://youtu.be/yO2yBBfqTDQ
• Dataset: http://bit.ly/datasetspss
