“Identifying and Understanding the Attributes Affecting Mobile Phone

Purchase”

Consumer Analytics Assignment Report

By Group 1:
Vaishali (404)
Ashirwad (406)
Prakhar (408)
Nishant (409)
 Table of Contents

1 Introduction 3

2 Research Objective 4

3 Research Methodology 4

4 Data Analysis and Findings 5 - 14

4.1 Multidimensional Scaling (MDS) 5-8

4.2 Discriminant Analysis (DA) 8 - 11

4.3 Perceptual Mapping 12 - 14

5 Summary and Conclusions 15

5.1 Findings 15

5.2 Managerial Implications 15 - 16

5.3 Limitations 16

5.4 Future Scope 16

6 Annexure 17

6.1 Questionnaire 1 17

6.2 Questionnaire 2 17

2
1. Introduction

This report tries to identify and understand the attributes that contributes into
selecting a mobile phone brand over others using the technique of Multidimensional
Scaling, Discriminant Analysis and Perceptual Mapping.

1.1 Multidimensional Scaling (MDS)

MDS is used to detect underlying dimensions that tell us similarity or dissimilarity
(distance on Euclidean plane) among different investigated objects (brands). It arranges the
objects in a space with set number of dimensions so that the distance among them is easily
observable. Objects that are similar are located closer to each other on the graph.

1.2 Discriminant Analysis

It is a statistical tool used to differentiate between groups of objects based on their
attributes. It is performed on data where the dependent variable (objects) is categorical and
independent variables (attributes) used to differentiate are interval in nature. A combination
of the independent variables is used to make the differentiation among different groups of
attributes.

1.3 Perceptual Mapping

It is a graphical technique used to explain the perceptions of customers about
products, brands, attributes, etc. It can also be used to show perception and sales of several
products in order to study the market segment and positioning of competing products
pictorially.
Perceptual Map displays the Image of brands or products in the minds of people whereas
Hybrid Maps displays both Image & appeal.

1.4 Research Questions

The research has two questionnaires, the first questionnaire deals with questions
regarding Multidimensional Scaling (MDS) and the second questionnaire has questions
about Discriminant Analysis (DA). For MDS, respondents are asked to rate 8 different mobile
brands namely Apple, Samsung, OnePlus, LG, Huawei, Sony, Asus and Google on a scale of
15. The rating is a pairwise process of marking the level of dissimilarities between two
brands. 1 being least dissimilar and 15 being most dissimilar.
For DA, questions were asked to rate 6 attributes (RAM, storage, battery backup,
processor, affordability, camera performance) for 3 brands – Samsung, Apple iPhone and
Oneplus, on a scale of 7, with 1 being Least Important and 7 being Most Important.
2. Research Objectives

1. To find the comparative evaluations (dissimilarity) of different brands of mobile phones

when the basis of comparison is unknown
2. To identify primary dimensions to explain the dissimilarity between different brands.
3. To obtain a perceptual map of the identified attributes and understand the attributes
affecting the judgment of the respondent for different mobile brands

3. Research Methodology

Sample size: 4 respondents, questionnaire distributed via emails

Type of sampling: Convenience Sampling

Type of data: Primary Data

Subjects: The respondents are the students of LBSIM studying in the second year of Research
and Business Analytics.

Software Used: IBM’s SPSS, MS Excel and MS PowerPoint

Procedure:

The responses obtained were analysed in SPSS using Multidimensional Scaling method and a
Discriminant Analysis to plot Perceptual Map in MS Excel.
4. Data Analysis and Findings

4.1. Multidimensional Scaling (MDS)

Brand iPhone Samsung Sony Oneplus Google Huawei LG Asus
iPhone 0.00
Samsung 4.50 0.00

Sony 7.50 8.50 0.00

Oneplus 6.00 10.75 5.00 0.00

Google 5.25 4.75 6.25 8.00 0.00

Huawei 8.50 9.25 6.25 5.25 9.25 0.00

LG 4.50 7.50 7.00 8.00 8.25 8.75 0.00

Asus 9.75 10.75 8.50 6.00 8.00 8.50 8.50 0.00

Table 4.1.1: Average pairwise rating of dissimilarity (1-15 scale)

4.1.1. 3-Dimension Scaling:

Figure 4.1.1.1: COORDINATES of Attributes

Clearly, from the above 3-D picture, it can be found that that:

 1st Dimension: Here, two groups can be identified. One group which is in ORANGE
box are iPhone, Samsung & Google and the other group is in BLUE oval are Oneplus,
Huawei & Asus. Hence, this 1st Dimension is being named as → COUNTRY of
ORIGIN as Oneplus, Huawei & Asus are China based mobile companies whereas
iPhone, Samsung & Google are from South Korea & USA.
 2nd Dimension: Here, only one group could be identified in BLUE boxes that consists
of iPhone, Sony & Oneplus. Hence, 2nd Dimension is being named as → UNIQUE
VALUE OFFERRING as iPhone, Sony and Oneplus each have some unique thing. Apple
has iOS, Sony has 4K Bravia display and Oneplus has Slider.
 3rd Dimension: Here, two groups can be identified. One in BROWN box is iPhone,
Huawei & Asus and the other group in GREEN oval are Samsung, Oneplus & Google.
Hence, 3rd Dimension is being named as → SALES as each category so formed as
unique set of sales reach.

Figure 4.1.1.2: 3-D MDS Plot

4.1.2. 2-Dimensional Scaling:

Figure 4.1.2.1: COORDNATES of Attributes

Clearly, from the above 2-D picture, it can be found that that:

 1st Dimension: Here, two groups can be identified. One group which is in BLUE box
are Oneplus, Huawei & Asus and the other group is in PINK oval are iPhone, Google
& LG. Hence, this 1st Dimension is being named as → COST-BENEFIT RATIO as each
of the group so formed in this category has nearly same Cost-Benefit Ratio.
 2nd Dimension: Here, three groups could be identified in BROWN boxes that consists
of iPhone & Sony then in GREEN oval consisting of Oneplus & LG and in PURPLE
hexagon consisting of Google & Asus. Hence, 2nd Dimension is being named as →
MISCELLANEOUS as nothing in specific can be concluded
 Figure 4.1.2.2: 2-D MDS Plot

4.2. Discriminant Analysis (DA)

Analysis Case Processing Summary
Unweighted Cases N Percent
Valid 12 100.0
Excluded Missing or out-of-range 0 .0
group codes
At least one missing 0 .0
discriminating variable
Both missing or out-of- 0 .0
range group codes and
at least one missing
discriminating variable
Total 0 .0
Total 12 100.0
Table 4.2.1: Analysis Case Processing Summary

The above table describes an overall summary. Complete dataset was divided into three
categories “Valid”, “Excluded” and “Total”. Here, all the 12 observations have been declared
as valid. A total of four respondents have rated for three brands (Samsung, Apple &
Oneplus) thus resulting in a total of 12 observations.
 Group Statistics
Valid N (listwise)
Brand Mean Std. Deviation Unweighted Weighted
Samsun RAM 5.50 1.000 4 4.000
g Affordability 4.25 1.708 4 4.000
Storage 5.00 1.826 4 4.000
Camera 6.50 1.000 4 4.000
Processor 6.00 .816 4 4.000
Battery_Backup 5.25 1.500 4 4.000
Apple RAM 4.50 3.000 4 4.000
Affordability 5.50 2.380 4 4.000
Storage 4.25 2.754 4 4.000
Camera 4.25 2.217 4 4.000
Processor 3.50 2.380 4 4.000
Battery_Backup 4.25 1.708 4 4.000
OnePlus RAM 4.75 2.217 4 4.000
Affordability 5.25 2.062 4 4.000
Storage 4.50 1.291 4 4.000
Camera 5.00 1.826 4 4.000
Processor 5.75 .500 4 4.000
Battery_Backup 3.75 2.062 4 4.000
Total RAM 4.92 2.065 12 12.000
Affordability 5.00 1.954 12 12.000
Storage 4.58 1.881 12 12.000
Camera 5.25 1.865 12 12.000
Processor 5.08 1.782 12 12.000
Battery_Backup 4.42 1.730 12 12.000
Table 4.2.2: Group Statistics

Here, each brand has been further distributed into six sub-categories/ six attributes namely-
RAM, Screen Size, Storage, Camera, Processor and Battery Backup. Finally, a descriptive
analysis was obtained. The total of 12 observations were found to be equally distributed
among Brands “Samsung”, “Apple” and “OnePlus”.
 RAM has an overall Mean value of 4.92 whereas the Mean RAM value of Samsung is
5.50, Mean RAM value of Apple is 4.50 and Mean RAM value of Oneplus is 4.75. It seems
that that the mean values of RAM are scattered across brands.
 Affordability has an overall Mean value of 5.00 whereas the Mean Affordability value of
Samsung is 4.25, Mean Affordability value of Apple is 5.50 and Mean Affordability value
 of Oneplus is 5.25. Mean Screen size of Affordability seems to be quite deviated from
the overall mean.
 Storage has an overall Mean value of 4.58 whereas the Mean Storage value of Samsung
is 5.00, Mean Storage value of Apple is 4.25 and Mean Storage value of Oneplus is 4.50.
It seems that that the Mean value of Samsung is quite far from the overall mean.
 Camera has an overall Mean value of 5.25 whereas the Mean Camera value of Samsung
is 6.50, Mean Camera value of Apple is 4.25 and Mean Camera value of Oneplus is 5.00.
It seems that that the Mean values of Samsung is quite higher than the overall mean
value.
 Processor has an overall Mean value of 5.08 whereas the Mean processor value of
Samsung is 6.00, Mean processor value of Apple is 3.50 and Mean processor value of
Oneplus is 5.75. It seems that that the Mean values of brands are quite far from the
mean value (both higher & lower than overall mean values)
 Battery Backup has an overall Mean value of 4.42 whereas the Mean Battery Backup
value of Samsung is 5.25, Mean Battery Backup value of Apple is 4.25 and Mean Battery
Backup value of Oneplus is 3.75. The mean value of brand Apple is very close to the
overall mean value.

Eigenvalues
Canonical
Function Eigenvalue % of Variance Cumulative % Correlation
a
1 1.906 82.8 82.8 .810
a
2 .396 17.2 100.0 .533
a. First 2 canonical discriminant functions were used in the analysis.
Table 4.2.3: Eigen value

In the above analysis, Two Functions (1 & 2) have been identified because we had taken
three discriminating variables (or) three types of brands as outcome of our discriminant
analysis function. Also, the number of functions is minimum of “One minus number of
groups/levels (in this case 3 for three brands)” and “The number of independent variables
(here 6). So, we ended up with Two Discriminating functions.
Eigen Value -A larger Eigen value explains a stronger function. The proportion of variance
explained by 1st Discriminant function is almost double and close to 2.0.
% of Variance – The first function alone is explaining a variance of 82.8 % and the second
function is explaining a variance of 17.2% only. The variance explained by 1st function is
almost 4 times as that of 2nd function. Hence, it can be assumed that 1st discriminating
function is
likely to discriminate among the brands with a decent accuracy. But we will be considering
both the discriminating functions for our discussion.
Canonical Correlation- A value closer to 1 indicates a better discriminating function. Here, 1st
discriminant function seems to be very good at discriminating.

Table 4.2.4: Test of Equality of Group Means

From the above table-4, it is found that Processor has the least value of “Wilks’ Lambda”
(0.566), it has the highest discriminating power. Rest other variables are not that good a
predictor. However, the p-value of processor is still around 7% (which is more than 5%)

Standardized Canonical Discriminant

Function Coefficients
Function
1 2
RAM .156 .421
Affordability -.296 -.305
Storage .410 .426
Camera .895 .464
Processor .907 -.737
Battery_Backup -.382 .718
Table 4.2.5: Standardized Discriminating Function Coefficients

Form the above table-5, a higher “Standardized Discriminating Coefficient” means the
variable has a higher discriminating power. Here, in this case: Camera (Std. Discriminating
Value = 0.907) has the highest discriminating power in Discriminant Function 1 and also in
Discriminant Function 2 (Std. Discriminating Value of Processor = -0.737).
Hence, it can be concluded that Processor can be the best predictor in predicting whether the
Brand will be “Samsung”, “Apple” or “Oneplus”.
Also, based on the above table, we can rank the all variables in terms of “Best Predictor of
Brand” with the following rankings:

DISCRIMINANT FUNCTION 1 DISCRIMINANT FUNCTION 2

1. Processor 1. Processor
2. Camera 2. Battery Back Up
3. Storage 3. Camera
4. Battery Backup 4. Storage
5. Affordability 5. RAM
6. RAM 6. Affordability

4.3 Perceptual Mapping

Canonical Discriminant Function

Coefficients
Function
1 2
RAM .070 .189
Affordability -.143 -.147
Storage .200 .208
Camera .510 .264
Processor .612 -.497
Battery_Backup -.216 .405
(Constant) -5.382 -1.792
Unstandardized coefficients
Table 4.3.1: Canonical Discriminant Function

On the basis of the above table-4.3.1, Equations will be formulated as:

DF_1 (Discriminant Function 1st): A +B1X1 +B2X2 + B3X3 + B4X4 + B5X5 + B6X6
DF_2 (Discriminant Function 2nd): C +D1X1 +D2X2 + D3X3 + D4X4 + D5X5 +
D6X6 Where:
A, C → Constant coefficients
B1, B2, B3, B4, B5, B6, D1, D2, D3, D4, D5, D6 → Coefficients for all the variables in
respective Discriminant Function formula.
X1, X2, X3, X4, X5 & X6 → These are the six variables
 Functions at Group Centroids

Function
Brand 1 2
Samsung 1.251 .519
Apple -1.611 .234
Oneplus .360 -.753
Unstandardized canonical discriminant functions
evaluated at group means
Table 4.3.2: Group Centroids for the Brands

From the above table-4.3.2, “GROUP-CENTROIDS” are found out. It will be plotted for the
Three Brands (Samsung, Apple, Oneplus) as shown in the following Figure to obtain a
“PERCEPTUAL MAP”.

The coordinates of Brands are:

[1] Samsung (1.251, 0.519), [2] Apple (-1.611,0.234), [3] Oneplus (0.360, -0.753)

The coordinates of attributes or variables are-

[1]RAM (0.070,0.189), [2] AFFORDABILITY (-0.143, -0.147), [3] STORAGE (0.200, 0.208), [4]
CAMERA (0.510,0.264), [5] PROCESSOR (0.612, -0.497) [6] BATTERY BACKUP (-0.216, 0.407)

Figure 4.3.2: PERCEPTUAL MAP FOR THE THREE BRANDS

From the above Figure-4.3.1 of “PERCEPTUAL-MAP”, following things have been found out:
1. Attributes:
1.1. RAM → Based on the survey, RAM is found to one of the defining criteria that
drives consumer’s buying behavior to the brand Samsung.
1.2. Affordability →Based on the survey, AFFORDABILITY doesn’t drive any of the
consumers towards the three brands considered in the Perceptual Map. Maybe, if
furthermore brands had been taken into consideration, AFFORDABILITY would have
been driving consumers to a certain brand.
1.3.Storage → Based on survey, STORAGE attribute drives the consumersto
purchase the brand of Samsung.
1.4.Camera → Based on survey, CAMERA attribute drives the consumersto
purchase the brand of Samsung.
1.5.Processor → Based on survey, PROCESSOR attribute drives the consumers to
purchase the mobile phones of brand Oneplus.
1.6.Battery Backup → Based on survey, BATTERY BACKUP attribute drives the
consumers to purchase the mobile phones of brand Oneplus.

2. Brands:
2.1. Samsung → Three attribute/variables that drives the consumers to purchase
Samsung mobile phones were found to be RAM, STORAGE & CAMERA. Among these
three, the location of Camera was found to be closest to the centroid of Samsung
thus indicating that Camera Quality is the highest driving force. Manufacturers of
Samsung mobile phones should consider keeping Camera of highest quality. Camera,
Storage & Ram has to be the Point of Differentiation for Samsung.
2.2. Apple → Battery Backup is the feature that manufacturers of Apple mobile
phones should consider as a Point of Differentiation. Hence, they should continue to
provide long lasting and light weight battery in their mobile phones.
2.3.Oneplus → The manufacturers of Oneplus has to provide high capacity
Processors to as Point of Differentiation. These processors are the biggest
driving forces to make consumer choose Oneplus over other Brands.
5. Summary and Conclusions
5.1. Findings
A total of 8 different mobile brands were taken for evaluation of dissimilarity among
them. The response received from the respondents revealed that two pair of brands namely
“Oneplus” & “Samsung” and “Asus” & “Samsung” had the highest dissimilarity among them.
The dissimilarity among them had a value of 10.75 each (average value of all responses
received). The next pair of brands that were highly dissimilar were “Asus” and “iPhone” with
the dissimilarity rating of 9.75 (average value of all responses received). Third most
dissimilar pair was “Huawei” and “Samsung” with the dissimilarity rating of 9.25 (average
value of all responses received). The LEAST dissimilar pair (or the most resembling pair) was
“Samsung” & “iPhone” with the dissimilarity rating of 4.50 (average value of all responses
received).
When the 8 mobile brands were compared for dissimilarity and were analysed using
3-D MDS (Multi Dimension Scaling), the three dimensions were identified as “Country of
Origin”, “Unique Value Offering” and “Sales”. When the same 8 brands were analysed in 2-D
MDS, the two dimensions were identified as “Cost-Benefit Ratio” and “Miscellaneous”.
A Perceptual Map that defines the image of a brand in the minds of the consumer
was plotted using Discriminant Analysis. It was found that RAM, Storage and Camera drives
a consumer to buy Samsung brand, Battery Backup drives the consumers to purchase
mobiles of Apple brand and Processors drive a consumer to buy mobiles of Oneplus brand.
5.2. Managerial Implications
5.2.1. Apple iPhone

Battery Backup

Apple’s iOS and hardware are very well optimised which results in a fantastic battery life even with a
physically smaller battery. Low battery Anxiety is a real phenomenon which is caused by dead phone
battery. Apple can capitalize on Android’s inability to optimize the battery efficiency.

5.2.2. Samsung

RAM

Samsung uses a proprietary user interface over the base Android OS but has always been able
provide a smooth and lag free experience. Samsung is also one of the few OEM who offer 12GB of
RAM on their flagships. They also have an edge over Apple in the multi-tasking department as
Samsung UI is fine tuned to handle multiple application windows at once. This is a unique feature
limited to only Samsung’s UI. They already market their Note series as a device that can do it all.

Camera

Samsung has been at the top of the camera tech for years now and shares the podium with Google
and iPhone. Like Apple, Samsung should market their phone’s camera prowess by putting up big
displays at various locations.
Storage

Among the big 3 – Samsung, Google and Apple, Samsung is the only one who gives base storage of
128GB and an option to further expand via micro SD card. Like Google markets that its customers
get unlimited cloud storage, Samsung can market they always give their customers the ability to
carry the unlimited storage with them. They could also exploit the vulnerability of data hack from
the cloud storage.

5.2.3. Oneplus

Processor

Oneplus has always used the top of the line silicon for their phones. Their tagline is “Never Settle”. A
new segment that is popping up is Gaming phones. Oneplus can easily target this segment as very
little changes such as – better heat dissipation, better CPU & GPU performance, etc. are required.
Their latest phone already has a must have gaming device feature- a higher refresh rate display.
Currently Gaming phone segment only has Xiaomi, Asus and Nubia; these are not known for their
premium quality, thus Oneplus has an advantage along with better brand recognition and bigger
market presence.

5.2.4. Affordability

 Since the brands taken in the study are premium brands with their phones selling for Rs.
50,000 or more, the affordability attribute does not point to any of the 3 selected brands.
Oneplus got its fame because of its incredible value proposition. But over the years Oneplus has
been increasing the price of its phones steadily, this might be a reason for it not getting the
affordability factor.

5.3 Limitations

1. The sample used is very small and might not be accurate in determining the
perceptions of the respondents about different brands of mobile phones.
2. Researchers’ beliefs and prior experience has been used to define the dimensions.
3. The attributes identified in the research may not be the exact attributes that have
been used by the respondents to evaluate
4. The results have not been tested by other methods
5. Evaluation of the brands are limited to the attributes provided

5.4 Future Scope

The sample used was very small and therefore, the future studies can be carried on
a large-scale sample. The respondents of this study were students, similar studies can be
carried out on different segments. Also, various statistical techniques can be clubbed
together to obtain a more accurate result in the future. A confirmatory analysis can be
carried out to check the reliability and validity of the research.
6. Annexure
6.1. Questionnaire 1
Q1-Q8. Please rate the level of dissimilarity between iPhone (Samsung, Sony, Oneplus,
Google, Huawei, LG & Asus) and following brands.
Please rate the following on a scale of 1 to 15. 1 being Least Dissimilar and 15 being Most
Dissimilar.

Least Most
Dissimilar 2 3 4 5 6 7 8 9 10 11 12 13 14 Dissimilar
-1 - 15
a b c d e f g h i j k l m n o
iPhone
Samsung
Sony
Oneplus
Google
Huawei
LG
Asus

6.2. Questionnaire 2
Q1-Q3. Please rate the level of Importance for each of the following features in a Samsung
(iPhone & Oneplus) phone.
Please rate the following on a scale of 1 to 7. 1 being Least Important and 7 being Most
Important.
Least Important - Most Important -
2 3 4 5 6
1 7
a b c d e f g
RAM
Affordability
Storage
Camera
Processor
Battery backup