CHAPTER THREE

METHODS

This chapter provides a comprehensive outline of the approaches the researcher took to address
the research goals put forth for the study; these regard the study design, the research
environment, the participants, the tools, the ethical considerations, and the statistical analysis.

3.1 Research design

A survey research design will likely be employed, using questionnaires distributed to students at
the University of Ibadan.

3.2 Setting
This study will be carried out at the University of Ibadan, in Ibadan, Oyo state. The University of
Ibadan is a federal government university that is located in the Ibadan North Local Government
Area of Oyo state, Nigeria. The reason why this setting was chosen is because the research aimed
to study effects of advertising and brand perception on consumer buying behaviour and the
researcher intends to use university undergraduate students as case study..

The University of Ibadan comprises of 16 faculties for undergraduate studies, namely: Faculty of
Agriculture, Faculty of Arts, Faculty of Basic Medical Sciences, Faculty of Clinical Science,
Faculty of Dentistry, Faculty of Economics and management sciences, Faculty of Education,
Faculty of Environmental Design and management, Faculty of Law, Faculty of Pharmacy,
Faculty of Public Health, Faculty of Renewable Resources, Faculty of Science, Faculty of the
Social Sciences, Faculty of Technology, and Faculty of Veterinary Medicine.

3.3 Sample Size Determination

The study made use of systematic sampling techniques. The reason why systematic sampling is
going to be used is because the researcher has the list of the population. Yamane (1967) formula
n= N/(1+N(e)2
n= the sample size

1
N = the population of the study
e = the margin error in the calculation
N= 1,500/(1+1,500(0.05)2
N= 1,500/(1+1,500(0.0025)
N= 1,500/(1+3.75)
N= 1,500/4.75
N= 315.

3.4 Participants
The study focused on University of Ibadan Undergraduate students and data will be obtained
from selected faculties in the university. The participants will consist of both male and female
students comprising 100 level, 200 level, 300 level, 400 and 500 level students will make up the
sample respectively.

3.5 Sampling technique

This research will adopt a convenience sampling. Data was gathered from respondents
(undergraduate students) in University of Ibadan, Ibadan.

6.5 Inclusion criteria

1. Participants must be undergraduate students of the University of Ibadan.
2. Participants must be students of one of the three selected faculties.

6.6 Instruments
A structured questionnaire consisting of four sections will be used in the study to measure the
impact of online advertising and consumer brand perception on online apparel consumer buying
behavior, researchers often employ various psychological scales and constructs:

Section A: Demographic Variables

Demographics were measured, including gender, age, education, occupation, ethnicity, marital
status and annual household income. Face validity of each item was checked by three apparel
and two marketing researchers.

2
Section B: Social Media Advertising
The validity of a measurement system is defined as the extent to which it accurately measures the
target phenomenon (Asika, 2006). Face validity, content validity, and construct validity were
utilised to evaluate the reliability and validity of the research instrument that was employed in
this study. In contrast, reliability calculates an estimate of the degree to which a measuring
device consistently produces the same results when employed on the same subjects or
components under the same circumstances. Therefore, the researcher used Cronbach's alpha, a
measure of internal consistency dimension, to assess dependability.

Table 3.1: Chronbach Reliability Analysis

Cronbach Alpha
Predicting/Criterion No. of Items
Variables Value
Social Media
5 0.821
Marketing
Patronage Intention 5 0.742
Consumer Purchase 5 0.764
Repeat Patronage 5 0.812

Section B: Brand perception (brand experience, brand image-congruence, brand affect, brand
trust, brand repurchase intent in online apparel shopping)
A survey questionnaire was used to measure five constructs: brand experience, brand image-
congruence, brand affect, brand trust, and brand repurchase intention. Questionnaire constructs
were designed based on previous researches. The items for measuring brand experience, brand
affect and brand repurchase intention were adapted from Chen-Yu et al. (2016). The items
measuring the variable brand image-congurence were adapted from Han (2006) and Chen-Yu et
al. (2016), and the items for measuring the brand trust variable were adapted from Park and Stoel
(2005) and Chen-Yu et al. (2016). All of the items were based on a 5-point Likert scale with the
range from stronlgy agree (5), agree (4), neutral (3), disagree (2) to strongly disagree (1).

Construct Cronbach’s CR Ave Adapted from

3
 α
Brand Experience 0.90 0.70 Chen-Yu et al. (2016); Bruner (2012); Yoon
0.899 &Youn (2016)
(BE)
Brand Image- 0.86 0.56 Chen-Yu et al. (2016); Han (2006)
Congruence (BIC) 0.87

Brand Affect (BA) 0.85 0.53 Chen-Yu et al. (2016); Bruner (2012);
0.825
Chaudhuri & Holbrook(2001)
Brand Trust (BT) 0.806 0.81 0.52 Chen-Yu et al. (2016); Park and Stoel (2005)
Brand Repurchase 0.88 0.64 Chen-Yu et al. (2016); Maxham III &
0.749 Netemeyer (2003); Tax et al., (1998)
Intention (RPI)

Section B: Consumer buying behavior

Validation of the EBBT scale
The results reported for the EBBT scale using the calibration data set may be upwardly biased
because they are based on the sample on which the instrument was developed. Therefore, the
measure was cross-validated on two separate samples of subjects from two different countries

Method. The EBBT scale was administered to a sample of 320 undergraduate students in a
program in hotel, restaurant, and institutional management at a major American university, and
to a sample of 159 undergraduate students at a university in the Netherlands. 4 The items were
scored as in the first study.

Results. The same two-factor structure was specified for the validation samples as for the
calibration sample, but all parameters were freely estimated (cf. Bagozzi and Yi, 1988). The
following model fits were obtained: ×2(169)=416.53 (p < 0.001), GFI = 0.88, CFI = 0.83 for the
American sample; and X 2 (169) = 260.15 (p < 0.001), GFI = 0.85, CFI = 0.88 for the Dutch
sample. 5 For both samples, GFI and CFI compare favorably with the results obtained in the
calibration sample, which supports the crosssample validity of the scale. Further, for the
American cross-validation sample, all factor loadings were significant (all t-values exceed 3.0)
and only three did not exceed 0.4, while in the Dutch sample, all loadings were significant (all t-
values exceeded 3.5), 15 exceeded 0.4 and all were greater than 0.3 in magnitude. The composite
reliabilities of EAP and EIS were 0.81 and 0.78, respectively, in the American sample, and 0.75

4
and 0.85, respectively, in the Dutch sample. The corresponding coefficient alphas were 0.80,
0.77, 0.74, and 0.85. In neither samples’ were any modification indices significant, and none of
the estimated changes exceeded 0.2.

Discussion. The results of the three scale development and validation studies show that it is
possible to construct an internally valid instrument for the measurement of exploratory buying
behavior tendencies based on our two-factor conceptualization of the construct. The findings
from three samples of consumers in two different countries indicate that the scale has good
psychometric properties, with a well-defined dimensional structure and adequate reliabilities.

Note: Items are listed by dimension (first 10 EAP items, then 10 EIS items); in actual
administrations of the scale, items should be listed in random order. Items are scored on 5-point
Likert scales, with 1 = strongly disagree and 5 = strongly agree
• Items marked with an asterisk are reverse scored•

3.7 Procedure

The survey aspect of study involved the researcher obtaining foremost ethical approval
by submitting a Research proposal to the Social Sciences and Humanities Ethics Committee
(SSHEC) and the University College Hospital (UCH), University of Ibadan. The researcher
administered the questionnaire among professionals at the University College Hospital (UCH),
University of Ibadan in the following fields: physicians, nurses, pharmacists, and medical
laboratory scientist. The administration was done via paper-based questionnaire.

3.8 Ethical Considerations

Informed consent will be obtained from participants, and data confidentiality will be maintained.
The research will adhere to ethical research guidelines.

3.9 Statistical Analyses

5
Descriptive statistics

Descriptive statistics was carried out to know the mean age, and the standard deviation in age,
ratio of male to female participants, distribution of participants across religion, distribution of
participants across different profession

Inferential statistics

Hypothesis 1 was tested using multiple regression analysis to analysis the degree to which
psycho-demographic variables will independently and collectively contribute to occupational
prestige.

Hypothesis 2, 3, 4 was tested using Pearson-r to find the nature and direction of relationship
between psycho-demographic variables and occupational prestige.

Hypothesis 5, 6 and 7 was tested using T-Test to find if there is a statistical significant difference
in responses of male and female respondents, respondents from low social economic status and
high social economic status.