We’re looking for a professional consultant who can help us determine how consumers evaluate our services.  How can ARI help us?  Please give an example. 

A regional bank from Albuquerque contacted ARI because senior management wanted to know how consumers in New Mexico evaluate banking services.  In other words, what banking attributes are important to consumers?  The insight from a recent focus group revealed several underlying dimensions but this was based on a small sample size and the bank wanted a more conclusive marketing analysis.  

The research objective for this project was centered around the statistical importance of various bank attributes so a survey was selected as the obvious method for collecting this information.  An online survey was recommended because it eliminated the costs associated with conventional survey methods, such as printing, postage and data entry.  In addition to these advantages, online surveys offer better response rates compared to questionnaires sent through the mail. 

For the next phase of the project, ARI developed a questionnaire to measure the importance of specific banking attributes.  For this abbreviated example, survey participants were asked to rate the importance of eight banking attributes on a 5-point scale ranging from unimportant to very important.  The banking attributes are shown below: 

V₁ How important to you is online banking?

V₂ How important to you is the interest rate on a loan?

V₃ How important to you is the obtainability of a loan?

V₄ How important to you are extended banking hours?

V₅ How important to you are auxiliary banking services?

V₆ How important to you are convenient ATM locations?

V₇ How important to you is a bank’s community involvement?

V₈ How important to you is the attractiveness of a bank’s exterior?

For this project the relevant population was defined as homeowners in six counties served by the bank between the ages of 32 and 65 with an estimated household income equal to or greater than $45,000.  In order to gain access to the targeted consumer group, ARI purchased a list from a trusted direct mail vendor.  This list was used as the sample frame for the project.

For the next phase of the project, ARI assisted the marketing director in drafting a survey invitation on the bank’s letterhead.  Participants were offered an incentive to participate with the assurance that their identity would not be disclosed to the bank.  Survey participation for any project is essential for the marketing research profession, therefore a participant’s anonymity and privacy must be protected.  The invitation directed participants to the ARI website which was used as the survey portal.

Sample size has a significant impact on the cost of any project, therefore ARI used a proven statistical formula to calculate the correct sample size based on estimated variance, the bank’s desired accuracy and the level of confidence needed for an estimate of the true population value.  Variance and accuracy are estimated with a percentage and the level of confidence is a value used to estimate the distribution characteristics of the sample.  These values enable us to scientifically determine the correct sample size for a survey.  For this project the correct net sample size was calculated to be 474 total participants with a precision level of ± 4.5%.

The selection method is a very important component of the research project because the selection technique, not the size of the sample, determines a sample’s representativeness.  In other words, everyone from the targeted population should have an equal probability of being selected into the sample, therefore a random sampling procedure was used to draw names for the survey invitations.  This sampling technique ensures an unbiased estimate of the true population value.  In spite of the best incentives, many people simply refuse to participate in surveys.  This situation is always an issue so the initial sample size was proportionally increased to 2,975 because of anticipated nonresponse.

An average response rate of 16% yielded 474 completed surveys within 10 days.  Data analysis was performed with advanced statistical software to obtain the information needed for the project’s research objective.  In other words, how do consumers from the targeted population  evaluate a bank? 

For this project, factor analysis was used to identify the underlying dimensions which consumers use to evaluate banks.  Factor analysis begins with the construction of a correlation matrix because we’re interested in the interdependent relationships among all the variables.  A partial matrix constructed from ten surveys is shown below.  (the actual matrix has 474 rows)  Cell values represent importance ratings for each of the eight attributes measured on a 5-point scale.

i.e. 1=Unimportant; 5=Very important 

RESPONDENT NUMBER	V1	V2	V3	V4	V5	V6	V7	V8
1	2	1	5	3	5	3	2	4
2	4	4	3	5	5	2	2	3
3	4	4	3	5	5	2	2	3
4	4	3	3	3	5	1	5	3
5	4	4	3	5	5	2	2	3
6	4	4	3	5	5	2	2	3
7	4	4	3	5	5	2	2	3
8	5	3	3	2	1	2	5	4
9	4	4	3	5	5	2	2	3
10	4	4	3	5	5	2	2	3

 Factor analysis uses a statistic based on a chi-square transformation to test the null hypotheses that the variables are uncorrelated in the population.  The chi-square value for the actual matrix was greater than the critical value of the test statistic.  This means that the variables within the matrix are correlated.  If correlations between the variables are not significant then conclusions based on factor analysis would be inappropriate. 

In order to summarize the data within the matrix, a smaller number of factors must be extracted.  Several statistical procedures are utilized to determine the resulting number of factors.  Retention of factors for this project was based on the amount of variance associated with each factor.  Factors with a variance less than 1 are not retained using this procedure.  This is done because ARI looks for meaningful dimensions within the matrix.  The variance of these factors is shown in the following table.  Focus your attention on the Total column. Only three factors are greater than 1, thus they were extracted for a closer analysis. 

	Initial Eigenvalues
Component	Total	% of Variance	Cumulative %
1	3.237	40.466	40.466
2	1.763	22.032	62.498
3	1.298	16.226	78.724
4	0.629	7.857	86.581
5	0.510	6.372	92.953
6	0.308	3.853	96.806
7	0.209	2.612	99.418
8	0.047	0.582	100.000

The final interpretation of factor analysis is facilitated by identifying the variables that have large loadings on the same factor.  Three factors were extracted based on the data from the table shown above.  Which banking attributes are more closely related to factor 1, 2 or 3?  Variables with large loadings on the same factor have something in common.  The factor matrix shown below offers answers to these questions.  The individual values shown in the numbered columns represent factor loadings or correlations between three factors and eight variables; V₁ through V₈.

	Factor
	1	2	3
How important to you is online banking?	0.171	0.880	0.316
How important to you is the interest rate on a loan?	0.934	0.195	-0.066
How important to you is the obtainability of a loan?	0.962	0.084	-0.124
How important to you are extended banking hours?	-0.027	0.834	0.373
How important to you is auxiliary banking services?	0.898	-0.047	0.224
How important to you are convenient ATM locations?	0.137	0.856	0.284
How important to you is a bank's community involvement?	-0.089	0.269	0.770
How important to you is the attractiveness of a bank's exterior?	0.226	-0.081	0.714

Factor 1 has high coefficients for auxiliary banking services, obtainability of loans and the interest rate on loans.  Therefore this factor was labeled traditional banking services.  Factor 2 has high coefficients for online banking services, extended banking hours and convenient ATM locations.  Therefore this factor was labeled convenience.  Factor 3 has high coefficients for the attractiveness of a bank’s exterior and community involvement.  Therefore this factor was labeled visibility.

All of the factors are significant however more attention is focused on Factor 1 because it explains 40% of the total variance.  Variability indicates how similar or dissimilar respondents are compared to an average value.  i.e. assessment of bank attributes.  Advantage Research concluded that consumers evaluate banks using three basic factors; traditional banking services, convenience and visibility.

 What you have just read is a brief example of applied marketing research.  The survey data shown in this matrix is fictitious.  ARI will never reveal the identity of a client or the actual results of a project under any circumstances without a client’s consent.  This example was developed to help potential clients fathom an application of factor analysis which might be useful to their profession.