This study examined factors affecting repurchase intention of fitness equipment. The researchers collected data on repurchase intention (dependent variable) and factors like consumer trust, gender, education, income, consumer satisfaction, and service quality (independent variables). Univariate analysis found most respondents were male, had a bachelor's degree, middle income. Bivariate analysis showed relationships between variables. Multiple regression identified the best predictors of repurchase intention as consumer satisfaction and service quality. Moderation and mediation models were tested to better understand these relationships. The results provide insight into what drives repeat purchases of fitness equipment.
1. Examination of Factors Affecting
Repurchase Intention of Fitness Equipment
By: Chris and Andrew
2. Background
• What:
• Fitness Equipment
• So What:
• Fitness (Sports) Equipment is expected to grow 6.18% per year over the next 5 years.
• Covid has changed how people exercise and has increased the use of home gyms (Martin,
Champ, and Franklin 2021).
• What’s New:
• Breakdown of detailed demographics affects on repurchase intention
4. Research Aims
• What factors will affect a consumers repurchase intention of fitness
equipment?
• What demographic variables influence fitness equipment purchase intention?
5. Topic: Examination of
Factors affecting
repurchase intention
of fitness equipment
Repurchase
Intention
[DV]
Consumer
Trust
[IV]
Gender
[IV]
Education
[IV]
Income
[IV]
Consumer
Satisfaction
[IV]
Service
Quality
[IV]
6. Variable classification Variable name Purpose
A. Continuous variables
(numerical)
1.Repurchase intention One sample t-test, Two-sample t-test,
ANOVA, Regression
2. Service Quality Regression
3. Consumer Trust
4. Consumer Satisfaction
7. B. Discrete variables
(categorical)
5. Gender Two-sample t-test, ANOVA, Regression
Male
Female
6. Education ANOVA, Regression
a) No education
b) High School
c) Bachelors
d) Masters
e) Ph.D.
7. Income ANOVA, Regression
a) Poor
b) Lower income
c) Middle income
d) Upper middle income
e) Upper
8. Selection Criteria and Data Manipulation
• Sample Selection Criteria:
• Unknown as we are using open-sourced pre-collected data.
• Data Manipulation
• Being samples have complete data sets, no manipulation such as list wise deletion will
occur.
• Data Transformation:
• Our data was approximately normally distributed.
9. Modeling approach
• Regression model: A predictive model designed to analyze
the relationship between independent and dependent
variables. We will use mostly linear modeling. This model
helps determine the relationship between variables,
forecasting, and modeling.
• Step Wise Approach
25. R-square is the proportion of variation in the dependent
variable accounted for by the factor variable. It is equivalent
to eta-square in a one-way ANOVA.
The F-test in the regression output is a test of the
equivalence of the group means, just as it is in the case of
one-way ANOVA.
The R-square in this model indicates that the dummy
variables representing Educational group account for 2.5%
of the variation in repurchase intention, which is a
statistically insignificant amount [F(4,363)=2.305, p=.058].
Education E1 E2 E3 E4
1 (No Education) 1 0 0 0
2 (High School) 0 1 0 0
3 (Bachelors) 0 0 1 0
4 (Masters) 0 0 0 1
5 (PhD) 0 0 0 0
Education – Dummy coding
26. 𝑦 = 𝑏0 + 𝑏1𝐸1 + 𝑏2𝐸2 + 𝑏3𝐸3 + 𝑏4𝐸4
𝑦 = 4.38 + 0.75𝐸1 + 0.33𝐸2 + 0.54𝐸3 + 0.19𝐸4
The intercept is the mean Repurchase Intention for individuals who are Ph.D.. The mean is 4.38
The slope for E1 is the difference in means between individuals in the No education and those who are Ph.D.. Thinking
about the coding of E1 (where 1=Edu_No), the slope for 𝑏1 indicates that individuals who have no Education scored 0.75
points higher on average than those who are with Ph.D.. In other words, a person who has no education is expected to
score approximately 0.75 points higher on repurchase intention than a person who is with Ph.D.. This difference is
statistically insignificant (p<.212).
𝑦 = 4.38 + 0.75(1) + 0.33(0) + 0.54(0) + 0.19(0)) = 4.38+ 0.75 = 5.13 (mean for persons who have no education)
E1
E2
E3
E4
27. 𝑦 = 𝑏0 + 𝑏1𝐸1 + 𝑏2𝐸2 + 𝑏3𝐸3 + 𝑏4𝐸4
𝑦 = 4.38 + 0.75𝐸1 + 0.33𝐸2 + 0.54𝐸3 + 0.19𝐸4
The intercept is the mean Repurchase Intention for individuals who are Phd. The mean is 4.38
𝑦 = 4.38 + 0.75(0) + 0.33(1) + 0.54(0) + 0.19(0) = 4.38+ 0.33 = 4.71 (mean for persons who are with High School
education)
𝑦 = 4.38 + 0.75(0) + 0.33(0) + 0.54(1) + 0.19(0) = 4.38+ 0.54 = 4.92 (mean for persons who are with Bachelor's
degree)
𝑦 = 4.38 + 0.75(0) + 0.33(0) + 0.54(0) + 0.19(1) = 4.38+ 0.19 = 4..57 (mean for persons who are with Master’s degree)
E1
E2
E3
E4
28. Comparison of results with one-way ANOVA with Education as
the independent variable and Repurchase Intention as the
dependent variable.
29. We were able to generate those means through substitution of values
for our dummy variables into the prediction equation.
𝑦 = 4.38 + 0.75𝐸1 + 0.33𝐸2 + 0.54𝐸3 + 0.19𝐸4
The mean for the Ph.D. (reference category) is equal to the
intercept: 4.38
The mean for persons who have no education: 𝑦 = 4.38 + 0.75(1) + 0.33(0) + 0.54(0) + 0.19(0)) = 4.38+ 0.75 = 5.13
The mean for persons who are with High School education: 𝑦 = 4.38 + 0.75(0) + 0.33(1) + 0.54(0) + 0.19(0) = 4.38+ 0.33
= 4.71
The mean for persons who are with Bachelor's degree: 𝑦 = 4.38 + 0.75(0) + 0.33(0) + 0.54(1) + 0.19(0) = 4.38+ 0.54 =
4.92
The mean for persons who are with Master’s degree: 𝑦 = 4.38 + 0.75(0) + 0.33(0) + 0.54(0) + 0.19(1) = 4.38+ 0.19 =
4..57
Estimated Marginal Means
30. Education E1 E2 E3 E4 E5
1 (No Education) 1 0 0 0 0
2 (High School) 0 1 0 0 0
3 (Bachelors) 0 0 1 0 0
4 (Masters) 0 0 0 1 0
5 (PhD) 0 0 0 0 1
𝑦 = 5.13 + (−.42)𝐸2 + (−.21)𝐸3 + (−.56)𝐸4 + (−.75)𝐸5
The intercept is the mean Repurchase Intention for individuals who are
with no education. The mean is 5.13
33. The R-square change represents the proportion of variation in the
dependent variable uniquely accounted for by ‘Education’ status and
the F-test is a test of the variable.
To test the effect of the compound variable Education, we use
hierarchical regression where we remove variables (rather than add
them to the model). We start with the full model (block 1) and remove
the dummy variables (in block 2).
34. Here is the output from an ANCOVA that was
run with Education as the IV and Income as
the covariate.
The F-test for the effect of Education is
equivalent to that shown in the model
summary table from our regression below.
36. Variable Explanation
• Consumer Trust is the propensity to believe “the word, promise, verbal
or written statement of another individual or group can be relied
upon” (Rotter 1967, p. 653)
37. Variable Explanation
• Service Quality involves three main dimensions (interaction, environment,
and outcome) and each has subdimensions. Furthermore, customers
aggregate their evaluations of the subdimensions to form their perceptions
of an organization's performance on each of the three primary
dimensions. Those perceptions then lead to an overall service quality
perception.
• Brady MK. Some New Thoughts on Conceptualizing Perceived Service
Quality: A Hierarchical Approach. Journal of Marketing. 2001;65(3):34-49.
doi:10.1509/jmkg.65.3.34.18334
38. Variable Explanation
• Customer satisfaction arises when consumers compare their perceptions
of the performance of a product or service to both their desires and
expectations. This comparison process produces not only feelings of
satisfaction with the product or service, but also feelings of satisfaction
with the information on which their expectations are based.
• Spreng RA, MacKenzie SB, Olshavsky RW. A reexamination of the
determinants of consumer satisfaction. Journal of Marketing. 1996;60(3):15.
doi:10.1177/002224299606000302
39. Variable Explanation
• Income is a gain or recurrent benefit usually measured in money that
derives from capital or labor (Merriam-Webster).
40. Variable Explanation
• Educational levels are defined by the developmental differences of
students and how the learning environments are structured.
• https://safesupportivelearning.ed.gov/training-technical-
assistance/education-level
41. Variable Explanation
• Gender is the behavioral, cultural, or psychological traits typically
associated with one sex (Merriam-Webster).
42. Variable Explanation
• Repurchase Intention - Determination to act in a certain way, in this
case to purchase products again from the same company
• Collins, W.A. (1980). Development of cognition, affect, and social
relations. Volume 13
46. The variance inflation factors (VIF) can be used to diagnose
multicollinearity. Values above 10 are indicators of strong
multicollinearity.
Absence of strong multicollinearity
47. Values above 15 can indicate multicollinearity problems, values above 30 are a very strong
sign for problems with multicollinearity.
If you find two or more values above .90 in one line you can assume that there is a
collinearity problem between those predictors. If only one predictor in a line has a value
above .90, this is not a sign for multicollinearity.