Service Quality Measurement - Data Management

Term Paper (Advanced seminar) 2003 36 Pages

Business economics - Operations Research


Contents :

i. Abstract

1.0 Introduction

2.0 Part A - Questionnaire “Palavrion Corporation”
2.1 Comments - Strengths & Weaknesses
2.2 Probability Based Sampling Strategy
2.3 Non-Probability Based Sampling Strategy

3.0 Part B: Service Quality Satisfaction Survey - ADMECO AG
3.1 Introduction
3.1.1 Overview of ADMECO AG
3.2 Problem Description & Objectives of Survey
3.3 Definition of Population & Sampling Process
3.3.1 Population
3.3.2 Sampling; Consideration & Process
3.4 Questionnaire Design
3.4.1 Procedure & Guiding Principle
3.4.2 Guiding Principle
3.4.3 Design & Layout
3.4.4 Pre-test the Questionnaire
3.5 Result of Survey
3.5.1 Response Rate
3.5.2 Overview of Answers
3.5.3 Personal Interaction - Analysis of Section
3.5.4 Business Savvy - Analysis of Section
3.5.5 Added Value - Analysis of Section
3.6 Conclusion

4.0 Part C: Middling Records - Data Analysis
4.1 Introduction
4.2 Presentation of Data
4.3 Retrospective Analysis
4.3.1 Model Validation Multiple R
4.3. 1.2 Correlation Coefficient (R ) Confidence Interval (CI) Outliers
4.4 Regression Model of CQSI - Forecasting 4 th Quarter
4.5 Comparison between Outlets - Impact on Sampling Strategy/Forecast

5.0 Part D - Reflection
5.1 Sampling Process/Questionnaire
5.2 Data Analysis

6.0 Appendices
6.1 Palavrion Corporation Survey Form
6.2 Service Quality Satisfaction Survey - ADMECO AG
6.3 Middling Records Data Tables

7.0 References

i. Abstract

Over the past decade Service Quality Measurement (SQM) has been receiving more attention worldwide and taking a more central role as a measurement of success. The notion of Service Quality is found well documented throughout the literature and describes the interactive process between the customer and the service provider. In general, the SQM is a powerful technique to monitor customer satisfaction, helping to focus on key areas of improvement in order to establish a new baseline to the current service quality rating.

The importance for service organisations is twofold: firstly, in the implications entailed in how to choose the most appropriate technique of service quality measurement, and secondly, in how service organisations can influence the perceptions of an individual customer/user in relation to the service encounter he/she is participating in. Empirically, this involves moving away from a standardised model of the same service for everybody to an approach where the best way of achieving excellent performance lies in addressing the subjective needs of an individual customer, providing precisely the specific service that reflects customers/ individuals’ perceptions.

In other words, service is not so much what the business does, per s e, but what the customer experiences (Martin 1999i).

1.0 Introduction

This paper is divided into four parts considering the practical and theoretical approach of data management.

Part A discusses a Palavrion Corporation questionnaire in terms of its strengths and weaknesses and how the probability based as well as the non-probability based sampling strategy could be applied. Part B considers the real approach adopted at ADMECO AG in questionnaire design and in the sampling process to determine staff satisfaction with the IT Service Provider PH Networks1. In Part C, I will analyse, and elaborate on, a survey performed at Middling Records, creating a forecast for a given period. Finally, last but not least, Part D offers reflections on my main learning points and the insights gained in this assignment.

2.0 Part A - Questionnaire “Palavrion Corporation”

In a move acknowledging how satisfied and dissatisfied customers affect the bottom line of a restaurant chain, the Palavrion Corporation management decided to make use of a questionnaire for data feedback. Their decision reflects my own view that the gap between satisfied and completely satisfied customers can be large enough to pose a serious threat to viable businesses - one key reason why successful organisations need to make satisfaction analysis an integral part of their business approach. Section A of the assignment discusses and analyses the questionnaire used by the “Palavrion Corporation”, judging the design, its strengths and weaknesses and how it could be applied in a range of cases.

Palavrion Corporation is a Canadian casual and fine dining restaurant chain in operation since the 1980s. As part of their customer awareness programme, they have designed a customer satisfaction questionnaire (CSQ) to determine how customers perceive and experience their visits to the restaurants. I will consider the questionnaire in terms of its strengths and weaknesses and detail how a different strategy could be utilised, i.e., probability based as well as non-probability based sampling.

2.1 Comments - Strengths & Weaknesses

The initial impression of the questionnaire is positive and appealing, with the added pluses of userfriendliness and simple feedback analysis. In my view, once such a form is completed it will provide an answer to the underlying question: “Are my customers satisfied?”.

The questionnaire is divided into three sections (see copy of questionnaire in appendices, section 6.1) and its strengths and weaknesses are judged as follow:

Abbildung in dieser Leseprobe nicht enthalten

However, in general terms, the survey lacks any incentive to complete the form!

2.2 Probability Based Sampling Strategy

Probability sampling utilises some form of a random selection method and ensures that each member of the population has a calculable, non-zero probability of being included in the sample (Belton, 2001, p. 48). The sample frame for the survey would use all its current and past guests. Using the full sample can eliminate any bias through poor sampling selection, simplifying the sample frame design and potentially contributing to increased accuracy. A typical source of error is a non-response with the questionnaire failing to identify major areas for improvement and past and current guests deciding not to respond due to a lack of incentives.

In the Palavrion Corporation case under consideration, it may well be advisable to apply a proportional stratified random sampling method (Gober, 2001iii). To increase the efficiency of the sampling design, we would use different sampling fractions reflecting the restaurant chain’s primary focus on customer’s age with division into categories (strata) and random sampling within those categories. In this instance, the strata are customer age with the key focus on the 25-40 age group. The waiter would have to take an active role as an observer in judging customer age, since this is hardly a question to be asked after the customer has completed the questionnaire. The groups could then be sampled disproportionately to their representation in the population by age; in other words, 35% for the category between 25-40, 25%: 41-60, 5%:>60, 25%; 12-24, 10%: <12. Given a sample size of 1000, a random sample would then be made in line with their representation within each group.

For children in the category below the age of <12, the restaurant should provide a table set and invite the youngsters to participate in a colouring competition, providing their address and birthday. The restaurant could then use this information to build a database for annual birthday card invites to the specific addresses, with a pre-set price discount for the child’s birthday. This should result in a continuous steady stream of reoccurring customers to the restaurant chain around a child’s birthday. To make this an effective tool, the waiter would need to be trained the present the questionnaire effectively and motivate the customer to fill it in when presenting the bill. One can assume that repeat customers may not be inclined to complete the questionnaire again without further incentives, for example, participating in a monthly draw for a free luncheon. The most likely channel used to reach the customers surveyed will be mail, since probability-based sampling entails a selection and there are no precise indications on the ways the author of the questionnaire intends to reach the targeted sample. On the negative side, though, it must be noted that the non-response rate tends to be higher for self- administrated surveys (Belton, 2001 p.54).

2.3 Non-Probability Based Sampling Strategy

Non-probability based sampling methods are generally less demanding but do not guarantee obtaining the unbiased selection of ind]ividuals generated by using probability sampling. Surveyors tend to favour random sampling methods (probabilistic) over non-probabilistic ones and consider them to be more precise. Non-probability sampling methods can be divided into accidental or purposive types; because we usually approach the sampling problem with a specific plan in mind, most sampling methods are purposive in nature (Trochim, 2000iv).

The restaurant chain could well adopt a non-probability based sampling strategy simply by using non- proportional quota sampling to select individuals non-randomly according to fixed quota, such as their age, in scales similar to those in section 2.2. Non-proportional quota sampling is less restrictive as compared to proportional sampling as it does not consider the number that match the proportion of the population. This is a non-probabilistic analogue of stratified random sampling, classically used to assure that smaller groups are sufficiently represented (Trochim, 2000).

The restaurant chain could make the questionnaire available in different locations, and have it made easily accessible with the menu at each table. In addition, the waiter should present the questionnaire with the bill and a short introductory request to kindly fill in the form. An incentive could be a free cup of coffee or a marketing article such as a ballpoint pen. Commonly, only extreme negative or positive opinions might be included in the sample - one possible disadvantage of choosing such a route. Of course, another option is to use the classical method of customer interviews by asking a few questions; this may or may not correctly represent the customers’ population depending on the point and duration of the sampling process. One further distortion may be caused by the waiters being unlikely to interview repeat guests more than once.

3.0 Part B: Service Quality Satisfaction Survey - ADMECO AG

Frequently, a customer satisfaction survey (CSS) is used to measure the rate of customer or client satisfaction. This survey aims to identify high opportunity areas for improvement in the internal business performance at ADMECO AG.

3.1 Introduction

The second part of the assignment requires designing a sampling process and questionnaire with subsequently analysis of the data collected. At present, I work in Business Development for NMS, an investment group in Denmark, which is the majority shareholder of ADMECO AG. A year ago, I was commissioned to streamline IT operations and outsourced this job to PH Network AG, a local IT service provider. The questionnaire is designed to assess aspects of staff IT service quality at ADMECO AG and uses a modern questionnaire design and sampling process to determine staff satisfaction with their IT service provider. The survey will be tested on a minimum of 10 respondents.

In order to understand the scope of the present customer satisfaction survey, I will first highlight background information about ADMECO AG and then go on to explain the reasoning behind the questionnaire and discuss how far the information gathered will be useful to the company.

I will further elaborate on the population, discuss recommendations for the sampling process, the design of the questionnaire and comment on the process of testing and revising the questionnaire. The last section will contain a statistical analysis of the results illustrating and translating the responses into a meaningful result.

3.1.1 Overview of ADMECO AG

ADMECO AG2 is a small medical device company offering concepts and solution for the hospital market, in particular, operating theatres and clean room environments. The company was founded in 1981 and employees 26 people. ADMECO AG is represented in 23 countries globally, relying on independent service distributors, who sell, install and service ADMECO’s products on an exclusive basis to the hospital operating room environment in their markets

3.2 Problem Description & Objectives of Survey

A year ago ADMECO’s dependence on its information technology (IT) infrastructure was increasing while the demand placed on it had grown beyond the infrastructure capabilities at that time. I was charged with the task of improving the situation and was face with the immediate challenge of investigating and outsourcing an acceptable and efficient quality IT service to everyone at the company. At that time, the connection from their home to up-link to the corporate server had to be able to support the demands of the technical applications they were using. Subsequently, in terms of delivery, the network service and operations and company network infrastructure was upgraded significantly to provide the capacity to speed up connectivity to desktops and other specific locations in buildings.

The core problem ADMECO faces today lies in utilising this new technology efficiently and communicating effectively throughout the organisation within set time frames. A year ago, when PH Networks was asked to upgrade IT and provide connectivity to all members of the staff, we were also faced with the dilemma of how to educate some staff members on simple IT usage as a whole.

The ADMECO questionnaire has been designed to collect information to evaluate staff satisfaction of their IT service provider PH Networks AG and to determine the extent to which the goals and objectives of ADMECO’s IT outsourcing strategy have been met. This involves examining the effectiveness and utility of the programme in its entirety and the success of individual performances in meeting their personal milestones and objectives.

3.3 Definition of Population & Sampling Process

3.3.1 Population

Since we are conducting a staff survey, the target population is obvious. This contrasts with situations where, for example, the aim is to determine the likely success of a product where the target population may be less apparent. Nonetheless, the general rule is still valid that determining the correct target population is critical to the success of the research and achieving meaningful results. If you are not sampling the right population, or interviewing relevant individuals, your goals cannot be successfully met.

In my case, since the sample population was clear, I next needed to decide on how many employees to contact; this meant considering factors such as time available and necessary degree of precision. The further key factor of budgetary concern was not, in this case, an issue.

3.3.2 Sampling; Consideration & Process

According to Webster’s dictionary3 the definition of “sample” is clear: A finite part of a statistical population whose properties are studied to gain information about the whole. The goal is to define the sample is such a way that the result will offer a representative picture of the underlying population. If we turn to the concrete case of ADMECO and the Service Quality Satisfaction survey, that the population is obvious allows clusters to be set up or permits strata sampling. For example, I would be able to orchestrate sampling according to the employee’s level of IT proficiency within a department categorised into groups to represent the stratum. For the purpose of the exercise I decided to narrow down the sample by using stratified random sampling, selecting at random on three strata (Group A ,B & C);departments; Administration/Technical, Sales / Marketing and Manufacturing / Production.

Abbildung in dieser Leseprobe nicht enthalten

Table 3-1 Sampling

Random sampling is sometimes also referred to as probability based sampling which utilises some form of random selection method, ensuring that each member of the population has a calculable, nonzero probability of being included in the sample (Belton, 2001, p.48). The advantage of random sampling is that the sampling error can be calculated. On the other hand, although non-probability based sampling methods are generally more straightforward, they don’t guarantee an unbiased selection of individuals in the way that probability sampling does.

3.4 Questionnaire Design

3.4.1 Procedure & Guiding Principle

My past experience in the company gave me a number of advantages in my approach to design of a questionnaire at my former employer. I was able to directly contact and involve other managers in the process and contact management at PH Network to optimise the questionnaire to produce the most significant results possible. The design involved a procedure where moving on to the next item depends upon the successful completion of the previous item, making it essential for respondents to complete each and every step.


1 PH Networks AG, Hochdorf, Switzerland - is an IT Network Solution company, est. 1981 & has currently 12 employees

2 ADMECO AG (www.admeco.com) est. in 1981, headquarter in Hochdorf, Switzerland, employees 26 people

3 2003 Merriam-Webster Dictionary, 2003, URL link: www.webster.com


ISBN (eBook)
ISBN (Book)
File size
1 MB
Catalog Number
Institution / College
University of Strathclyde
service quality measurement data management



Title: Service Quality Measurement - Data Management