Task 2.3 Record and collect relevant data where appropriate
Before even consider to you record the collect data, it is advisable to understand the type of data so as to be able to program into SPSS smoothly.
There are basically three types of data:
Ordinal data
Ordinal data need not involve measurement. It relates only to the order in which the data is classified.
For example, in UK academic awards, a good student will be classed as First. Next will be an Upper Second. Then a Lower Second and so on. Not all Firsts will have the same marks; only they pass a certain limit for that classification.
Nominal data
Nominal data has no ranking/ordering associated with it.
For example: city of birth; gender; etc.
Interval data
Interval data is like ordinal except we say the intervals between each values are equally split.
For example, in temperature, the difference between 20 and 30 degrees is the same magnitude as 78 and 79.
https://answers.yahoo.com/question/index?qid=20081203215352AAg56gC
A clear picture of this importance in SPSS is depicted as follow:
Figure 1 Types of data in SPSS
The next important item to consider prior to entering the data into the SPSS is the objectives of data analysis. This according to Sekaran (2003) are to get a ‘feel of data’, test the ‘goodness of data’ and hypothesis testing as depicted in Figure 2 below:
Figure 2 Flow diagram of Data Analysis Process
Getting data ready for analysis
Editing Data
The first task needed to be done is to edit the data. This means we have to decipher and code the data prior to its entering into the computer. For example, open end question which could be basically responded by “yes” or “no” may not be done clearly with follow up notes. This often is unstructured questionnaires require considerable amount of time to organise and manage the data. This task has to be done with great clarity.
There are occasions whereby respondent inadvertently or deliberately did not answer sensitive questions, such age, income, qualification and etc.. This has to be cross checked with other variables responded by the same person. For example, it is not possible for an employee to have a Phd if he is only 20 years of age; or a person having an income of more than $10,000 if he/she just has an ‘O’ level of qualification. In this situation the respondents must be recalled for clarification.
Blank returnIf it is more than 25%, the whole interview/questionnaires process has to be redo. If less than 5% and if the error is random then the next handling process would has to take place. In order to qualify for a random status the p value (Sig) must be < 0.05. In our actual case, it is a random data as it is > 0.05. hence, this EM method can be smartly used
Figure 3 – Checking Blank Data is a random data
Handling Blank Data
In our situation, we use SPSS to do an “Expectation Manipulation” (EM). This is done by:
Analysis; Missing Values; Key in selected Variables; Descriptive; T Test; Probability in the Table; Continue; EM; Normal; Create File Name; New; Continue; Ok. Open New file:
Figure 4 – Using SPSS to handle the Blank Data
Coding Data
Before data entry, the data has to be coded so that the SPSS can recognise and handle it efficiently and effectively. In our case:
Figure 5 - Coding of Gender; 1 = Male; 2 = Female
Categorising Data
We use SPSS as these commands: Transform; Recode into same variable to create our conversion.
Figure 6 – Using SPSS to categorise data
Creating for SPSS Programming
Direct entry into SPSS as below:
Figure 7 – SPSS Variable view of Data entry
FEELING OF DATA
After the completed questionnaires were returned, the next stage was to analyse the data to test the 8 hypotheses. According to Sekaran (2003), the 3 objectives of data analysis are to get a ‘feel of data’, test the ‘goodness of data’ and hypothesis testing.
Based on Figure 9, the most active variance is in the area of Staff Pay followed by StaffWork/Balance and then Job itself. This is a preliminary round of checking.
GOOD OF DATA
Figure 10 Correlation Checking of Customer Satisfaction
In Figure 10 Correlation Checking of Customer Satisfaction - checking resulted high correlation in repurchase intention and value respectively. A more detailed checking involves using multi-regression checking in the SPSS.
Figure 11 Multi Regression Checking of Customer Satisfaction
Figure 11 - using the regression method confirmed Value variable as the highest and final preference and independent variable of the customer in term of its correlation with customer satisfaction dependent variable.
As shown in Figure 12 and Figure 13 below, Job Itself, Company Environment and Staff Pay ranked respectively in position 1, 2 and 3. However, multi regression checking, only to Job itself as the key variable for action. Looking at the “Intention to Resign” column in both the correction and multi regression checking there is a negative correlation. We can take it as when staff satisfaction is high or when they are more satisfied with their job, they are less likely to leave the organisation.
Figure 12 Correlation Checking of Staff Satisfaction
Figure 13 Multi Regression Checking of Staff Satisfaction
Our Respondent’s (Customer’s) Profile
Chart 1 Age Group
Chart 2 Qualification
Chart 3 Gender
Chart 4 Salary Group In sum, it would appear that the respondents are mainly from the “21 – 30” age group, with ‘O’ level qualification and mainly Male.
HYPOTHESIS OF DATA
Checking the Hypothesis with 1 sample T test
Figure 14 Customer Satisfaction and Gender
Looking at Figure 14 above, there is no evidence to prove that there are difference in Customer Satisfaction Level between both gender as the Sig. (2-tailed) = > p value of 0.05. Hence we reject Ho and accept Ha.
Interpretation of Results/Conclusion
All the analysis by SPSS points to a direction of Customer Satisfaction involves the preferences of value as the key to customer’s satisfaction. In term of Staff Satisfaction, it is the Job itself that matters. These data give lead us to recommend improvement in Job Design as the key to satisfy the staff. Looking at the customer’s profile convince us that they are mainly the young respondents with ‘O’ level qualification. And value is their main concern.
The hypothesis of a concern that Gender would be a main concern is addressed when the 1 sample t test confirmed this Ho of different in evaluating Customer Satisfaction. Ha is the result denoting no evidence to support there is a difference in opinion between the Male and Female in this aspect.
Recommendations
* Since customer is concerned about value, the company must look into their pricing strategies. * Introduce a Loyalty program * Club discount * A concrete Customer Relationship Management program * Job rotation, Job enlargement and Job enhancement as motivation factor to drive staff’s commitments and productivity.
Limitations
From the result it would appear that the data is skewed towards younger respondents, a more even mix of sample should be recommended. For example, the assistance research should approach certain number of older customers, e.g.; divide equally into 3 categories of age group.
This research use Likert Scale in the Questionnaires. Whilst, this method is a simple and a logical way of data collection, it also has its disadvantage of overstating its correlation.
Over generalising the respondents in term of age group selection for the sample is a weakness by itself. From the data collected, the sampling of the on site and random respondents are skewed towards the younger ones. This must be addressed in future research.
Customer satisfaction which is linked to Customer Loyalty is an evolvement process. This research due to its short time frame and budget is unable to address and capture such dynamism.
The 8 independent variables identified from the literature review seemed in order and correlated with the dependent variables (Customer Satisfaction). However, in reality, to operate in tandem is a difficult task. Hence, other independent variables should be introduced to make the research a more homogenous one. Such variable could be ambience, room temperature, room quality, room colour, leisure/recreation facilities, business centre and etc.
