Final Exam: Analysis of Reliability, Correlation, and Variance

The final exam questions focus on evaluating the reliability of measurement scales using split-half and test-retest methods, determining correlations between variables using Pearson coefficients, and interpreting ANOVA results for model fitting. Question 5 discusses split-half reliability with a Cronbach's alpha result indicating high consistency. Question 6 explores test-retest reliability with a lower alpha value suggesting less reliable scales. Questions 7 and 8 analyze different correlations among satisfaction, performance, and brand loyalty variables using Pearson correlation. Questions 9 through 11 apply ANOVA to assess the influence of various factors on customer satisfaction, intention to revisit, and product preference, emphasizing model fit and significance. Question 12 critiques study methodologies for potential biases, while questions 13 and 14 discuss limitations in representativeness and measurement approaches.

FINAL EXAM QUESTIONS
SECTION TWO
Question 5
a) Split half reliability analysis refers to the extent to which a scale gives the same result
even after multiple repetitions. The data is dividing into halves thus reliability in the two
halves. The output is alpha 0.9051 refers to extent to which its consistent measure of
concept.
b) Cronbach’s test to determine if there existed biasness in splitting.
c) The alpha 0.9051>0.7 thus high reliability
Question 6
a) Test- Retest is to construct a reliable measurement scale, to improve the scales and finally
evaluate the scale in use.
b) Calculate the coefficient alpha which will identify the reliability of the scales.
c) The reliability coefficient 0.6559 <0.7 thus the scale are less reliable.
Question 7
a) Pearson correlation measures linear correlation between two variables. When output is 1,
positive linear correlation. When output is 0, no linear correlations while -1 indicates a
negative linear correlation.
b) Correlation between satisfaction and unreliability is -.5 thus weak negative correlation,
intent to buy .764 thus positive correlation and friendly 0.18 hence weak positive
correlation.
Correlation between unreliability and intent to buy and friendly -.35 and 0.5 are weak
negative and positive correlations respectively.
Correlation between intent to buy and friendly are 0.007 which is weak positively
correlated.
Correlation between satisfaction and friendly is weakly positively correlated.
c) The output is logical since all these values are between 1 and -1. They are within the
range.
d) The unreliability of the supply company gave dissatisfaction to the buying company.
Question 8
a) Pearson correlation measures linear correlation between two variables. When output is 1,
positive linear correlation. When output is 0, no linear correlations while -1 indicates a
negative linear correlation.
b) Likelihood to purchase competitors brand and product performance are negatively
correlated. Purchasing competitors brand lowers performance of our product.

Likelihood to purchase competitors and intention to purchase our brand is negatively
correlated.
Likelihood to purchase competitors and strength of our product name is weakly
correlated.
Product performance and intention to buy our product are positively correlated since
performance is due to intention to buy.
Product performance and strength of brand name are positively correlated.
Correlation between intention to purchase our brand and strength of brand name are
positively correlated.
c) Despite the competitor the brand has likelihood of picking hence marketing should be
done to cater for strength of brand name.
Question 9
a) Analysis of variance. The F statistic, R squared and coefficients of model. R squared
indicates that 77% the model fits, F value 58.296 is statistically significant and constant
being negative implies that the company to get output needs to get into their pockets
b) R squared indicates that 77% the model fits, F value 58.296 is statistically significant
while coefficient of the regression model -.22 indicates customers had difficulties to use
the product, 55% declare that the quality was upheld, the brand name though positive
contributes but in a small percentage, low prices , variety of colours and money back
guarantee positively contributed to model.
c) Customer satisfaction was positively contributed by the variety of colours, quality, brand
name, low prices and money back guarantee. The company should ensure they maintain
these to improve sales. Marketing should be implemented to improve brand name in the
market.
Question 10
a) Analysis of variance. The F statistic, R squared and coefficients of the model. The R
squared is 63.5% fits the regression model. The F statistics 70.165 by level of
significance it is statistically significant.
b) The model dependent variable intension to revisit café’ X the independent variables 21%
of customer liked the décor, quality of food negatively influenced the revisit by33%, 11%
influenced by music played in the café’, quality of coffee influenced most of the
customers while uniform influenced negatively the clients.
c) The cafe’ X should improve the quality of food by employing experts in cooking and
make more impressive uniform for staff.
Question 11
a) Analysis of variance. The F statistics, R squared and coefficients of the model. The
adjusted R squared is77.1 which is 77.1% fits the model that preference was influenced

by independent variables. The F statistics 58.289 given 0.000 is less than level of
significance hence statistically significant in the model.
b) The coefficients in the regression model of preference product preference, advertising the
product, low prices, range of size, money back guarantee positively influenced the buying
of product in different towns the difficulties in opening affected negatively preference.
c) The advertising of product should be improved to allow more sales. The branding
company should come up with a container that is easier to open.
Question 12
The study use a focus group cannot be used to represent the Australian population. A survey that
allows sampling and randomness could have been used. The focus group means the researcher
was predetermined thus biasness.
Question 13
a) A focused group is not a representation of the population. The number of stratus which is
3 is also small with groups of 10 to represent the population it’s biased.
b) The findings from these study show that these study was biased and cannot be used to
make analysis or predictions.
Question 14
a) Multi-item measure is designed to measure the respondent’s attitude towards more than
one attribute related to the stimulus object.
b) An example is deaths that occur in Australia i.e. roads accidents, suicides, killings,
assassinations etc
c) Predetermined factors are known by single questions. Maintaining status quo thus lack
frexibility.