INCORPORATING PROSPECT THEORY INTO DECISION-MAKING …ijeronline.com/documents/volumes/2016/Nov - Dec 16... · INCORPORATING PROSPECT THEORY INTO DECISION-MAKING FOR CUSTOMER SATISFACTION

INCORPORATING PROSPECT THEORY INTO DECISION-MAKING

FOR CUSTOMER SATISFACTION

Yu-Cheng Lee1, Feng-Han Lin2*, Cheng-Fu Hsiao3 1Department of Technology Management, Chung Hua University, No.707, Sec.2,

WuFu Rd., Hsinchu 300,Taiwan

E-mail address: [email protected] 2Ph.D. Program of Technology Management, Chung Hua University, No.707, Sec.2,

WuFu Rd., Hsinchu 300,Taiwan

E-mail address: [email protected] 3Department of Hospitality Management, Hsing Wu University, No. 101, Sec.1,

Fenliao Rd., LinKou District, New Taipei City 24452, Taiwan

E-mail address: [email protected]

*Corresponding author

Feng-Han Lin

Ph.D. Program of Technology Management, Chung Hua University, No.707, Sec.2,

WuFu Rd., Hsinchu 300, Taiwan. Tel: +886-932781043; Fax: +886-35726580

E-mail address: [email protected]

ABSTRACT

In the customer-oriented trend of the last three decades, customer satisfaction has increasingly

become the object of study. Although studies have adopted the Kano model and

importance-performance analysis (IPA) to explore the influence of quality attributes on customer

satisfaction, the role of consumer behaviour has not be considered in this association. The present

study proposes a novel methodology for identifying quality attributes by exploring customer

requirements, and focuses on effectively evaluating customer satisfaction by employing prospect

theory. This methodology integrates the quantitative Kano model and dual IPA to identify asymmetric

and nonlinear relationships between attribute-level performance and customer satisfaction. In addition,

the incorporation of prospect theory assists users in effectively evaluating satisfaction. Finally, a case

study is presented in this paper to highlight the management implications. The current findings

contradict the reasoning of the Kano model and IPA. The novel methodology provides quality attributes

categorisation to precisely fulfil customer needs and help managers make strategic decisions to meet

these needs.

KEYWORDS: Customer satisfaction, Prospect theory, Kano model, Importance-performance analysis,

Quality attribute

Yu-Cheng Lee, Feng-Han Lin, Cheng-Fu Hsiao, Int. Eco. Res, 2016, v7i6, 26 – 38 ISSN:2229-6158

IJER – NOV – DEC 2016 Available [email protected]

26

INTRODUCTION

In the highly competitive global market, a product or service must fulfil meet customer

needs. Several current studies are increasingly focusing on customer satisfaction.

The Kano model (Kano, Seraku, Takahashi, & Tsuji, 1984) has been widely used to

satisfy customer needs by identifying and classifying the quality attributers; however,

it cannot provide decision-making support because the Kano category remains

qualitative (Riviere, Monrozier, Rogeaux, Pages, & Saporta, 2006; Xu, Jiao, Yang,

Helander, Khalid, & Opperud, 2009; Shahin, Pourhamidi, Antony & Hyun Park, 2013).

Attribute performance and importance, obtained through importance-performance

analysis (IPA)—originally introduced by Martilla and James (1977) —can facilitate

decision-making (Ting & Chen, 2002; Matzler, Bailom, Hinterhuber, Renzl, & Pichler,

2004; Matzler & Renzl, 2007). However, the relationship between attribute

performance and importance is asymmetric; this problem can be resolved using few

sustainable solutions.

Several related studies have considered the positive potential of integrating the Kano

model with IPA to accurately identify key quality elements and explore customer

satisfaction (Kuo, Chen, & Deng, 2012; Yin, Cao, Huang & Cao, 2016; Meng, Jiang, &

Bian, 2015). However, these studies have not considered the role of consumer

behaviour in customer satisfaction. To fill this gap in behavioural economics

knowledge, the present study incorporated prospect theory in customer satisfaction to

predict consumer behaviour.

The purpose of this study was to examine customer satisfaction with regard to

psychology through a behavioural economics analysis. The specific aims of this study

were (a) to identify the impact of attribute-level performance on customer satisfaction

and (b) to elucidate the concept of prospect theory to construct a behavioural

decision-making model. Although the current research type remains in its infancy, this

study may aid in further predicting customer satisfaction. Moreover, compared with

the conventional method, the current alternative method should make

decision-making model much more desirable. Finally, this study was concluded with a

case study of bicycle user satisfaction to demonstrate the current method.



27

LITERATURE REVIEW

Kano Model

The Kano model (Kano et al., 1984) was developed as a two-dimensional quality

model based on Herzberg’s motivation-hygiene theory (1959). This model classifies

quality attributes into five categories on the basis of their impact on customer

satisfaction: must-be, one-dimensional, attractive, indifferent, and reverse quality.

In recent years, several studies have described the use of the Kano model during

product development and design (Kano, 2000; Atlason, Oddsson & Unnthorsson,

2014; Dominici, Roblek, Abbate, & Tani, 2016). Furthermore, several studies have

examined the relationship between attribute-level performance and customer

satisfaction (Mittal, Ross, & Baldasare, 1998; Tontini, Søilen, & Silveira, 2013). Tontini

et al. (2013) suggested that before attractive quality attributes are improved,

improvement of must-be quality attributes or one-dimensional quality attributes

should be prioritised.

IPA

IPA uses a two-dimensional matrix, based on the customer-perceived importance of

quality attributes and attribute performance, to develop marketing strategies (Martilla

& James, 1977). The matrix is divided into four quadrants by using attribute

performance and importance:

(1) Keep Up the Good Work: In this area, quality attributes are highly regarded by

customers and the provided performance is satisfactory. IPA suggests that the quality

attributes in this quadrant should be maintained.

(2) Concentrate Here: Customers value the importance of the quality attributes

beyond the provided performance; therefore, customers are not satisfied with the

quality attributes. Quality attributes in this quadrant should be improved first.

(3) Low Priority: Customers think that the quality attributes in this area are not of great

worth. Nevertheless, owing to low performance customers are not satisfied with the

quality attributes. These quality attributes should be improved, but with low priority.

(4) Potential Overkill: Customers perceive that the quality attribute performance is

high and are satisfied with all attributes; hence, these quality attributes are of low

importance.



28

Matzler et al. (2004) developed both IPAs for both satisfied and dissatisfied

customers to estimate the relative impact of each attribute for high and low

performance. Kuo et al. (2012) proposed the IPA–Kano model to provide specific

strategies for definite attributes; however, this model only classified more quality

attribute types through IPA.

Prospect Theory

Kahneman and Tversky (1979) proposed the concept of a value function and

established the psychological codes of gain and loss. The three basic principles of the

theory are as follows (Tversky & Kahneman, 1981):

1. Most people tend to avoid risk when facing “gain.”

2. Most people tend to prefer risk when facing “loss.”

3. People are more sensitive to loss than to gain.

Thaler (1985) proposed the model of consumer behaviour by using the prospect

theory value function. Transaction utility uses a hybrid of cognitive psychology and

microeconomics to evaluate purchases. This model derives compound outcomes of

gains and losses, which are used to evaluate single outcomes with a value function.

The outcomes of prospect theory could then be valued jointly or separately to yield

greater utility. They are valued jointly as )( yxv and separately as )()( yvxv . Four

possible combinations can be considered for the joint outcome: multiple gains,

multiple losses, mixed gain, and mixed loss. The value function of the prospect theory

was denoted as )(v . Segregation is preferred in multiple gains, )()()( yxvyvxv ,

whereas integration is preferred in multiple losses, )()()( yxvyvxv . Mixed gain,

)()()( yxvyvxv , involving net gain, prefers integration, whereas mixed loss,

)()()( yxvyvxv , involving loss gain, prefers segregation.

Several authors (e.g., Mittal et al., 1998; Ting & Chen, 2002) have hypothesised that

the S-shaped value function curve of prospect theory is applicable to customer

satisfaction. Mittal et al. (1998) indicated that the impact of negative attribute

performance on customer satisfaction is greater than that of the positive attribute

performance; moreover, utility-preserving attributes are applicable to the S-shaped

value function curve of prospect theory. Thus, prospect theory may be useful for

constructing a decision-making model of satisfaction utility.



29

METHODOLOGY

This study performed a survey to identify asymmetric and nonlinear relationships

between attribute-level performance and customer satisfaction. As reported

previously (e.g., Mittal et al.,1998; Matlzer et al., 2004), regression analysis with

dummy variables was applied to identify the (asymmetric) impact of attribute-level

performance on (overall) customer satisfaction to indicate attribute importance. The

Kano model and IPA were employed to detect the differences among the attributes.

Finally, prospect theory was applied to assist managers in decision-making.

Therefore, the steps of the methodology are given as follows:

Step 1: Identify the impact of attribute-level performance on customer satisfaction

Attribute performance ratings are recoded to yield the dummy variables such as

‘high-level performance’ (1,0), ‘average performance’ (0,0), and ‘low-level

performance’ (0,1). A regression analysis is conducted to estimate the impact of

different level performance on customer satisfaction. The high-level performance

regression coefficient is used to measure the importance degree of satisfaction,

whereas the low-level performance regression coefficient is applied to measure the

importance degree of dissatisfaction.

Step 2: Categorise the quality attributes on the basis of the Kano model

The quality attributes of a product or service are categorised on the basis of the

intensity of the impact on customer satisfaction, indicated by the regression

coefficient, namely importance.

Step 3: Construct a pair of IPA matrices

A pair of IPA matrices is constructed using performance and paring importance. Both

IPA for satisfaction and dissatisfaction are integrated to evaluate quality attributes that

are crucial to customers and require improvement.

Step 4: Market strategic decision-making

Different attributes vary in their impact on satisfaction. The level of importance is

served as a gain or loss. The Kano model and IPA are integrated to strategically

analyse and use prospect theory to make decisions.

The current methodology considers customer behaviour to suggest how decision

makers should prioritise resource allocation.



30

CASE STUDY

Statistical Analysis and Classifications

The global bicycle market has been growing in recent years. Here the

aforementioned methodology was applied to the function requirements of a bicycle for

sports. As suggested by an expert, eight questions were presented in the

questionnaires: Q1 = appearance, Q2 = colour, Q3 = cushion comfort, Q4 = brake

system, Q5 = shift system, Q6 = wheel set and transmission, Q7 = weight, and Q8 =

accessories. These questionnaires were then used to measure attribute performance

and overall satisfaction. The quality attribute performance was evaluated on a scale

of 1 (extremely low) to 9 (extremely high) and overall satisfaction on a scale of 1–100.

Data from 192 respondents were collected through face-to-face interviews.

The impact of attribute-level performance on satisfaction was next evaluated using a

regression analysis with dummy variables (Table 1). The two regression coefficients

served as the importance of attribute-level performance: regression coefficients of

high- and low-level performance were used to measure the degree of importance of

fulfilling and unfulfilling quality attributes, respectively.

Table 1 Impact of attribute-level performance on customer satisfaction

Item Regression coefficients Quality attribute

category High performance Low performance

Q1 appearance 0.316*** -0.288*** one-dimensional

Q2 color 0.326* -0.178*** one-dimensional

Q3 cushion comfort 0.253(ns) -0.103* must-be

Q4 brake system 0.232*** -0.286** one-dimensional

Q5 shift system 0.270*** -0.296*** one-dimensional

Q6 wheel set and transmission 0.188*** -0.381** one-dimensional

Q7 weight 0.167(ns) -0.217(ns) indifferent

Q8 accessories 0.152** -0.280(ns) attractive

ns = not significant. * P < .05. ** P < .01. *** P < .001.

As shown in Table 1, for the Q8 regression coefficients, high performance was

obviously more significant than low performance. If Q8 was adequate, customers

were satisfied; however, inadequate Q8 did not significantly influence customer

dissatisfaction, implying that Q8 is an attractive quality attribute. By contrast, for the

Q3 regression coefficients, low performance was more significant than high

performance. Therefore, if Q3 was inadequate, customers were dissatisfied;



31

adequate or more favourable Q3 could not satisfy customers either. In other words,

Q3 was a must-be quality attribute. The Q1 regression coefficients of high and low

performance were significant; therefore, if Q1 was adequate or more favourable,

customers were satisfied; otherwise, they were dissatisfied. In other words, Q1 was a

one-dimensional quality attribute. Similarly, Q2, Q4, Q5, and Q6 were determined to

one-dimensional quality attributes. As for Q7, the regression coefficients of high and

low performance were nonsignificant; hence, Q7 was considered an indifferent quality

attribute.

IPA for Customer Satisfaction

The samples were divided according to high and low average overall customer

satisfaction. The average of high satisfaction was high performance for satisfaction;

by contrast, the average of low satisfaction was low performance for dissatisfaction

(Tables 2 and 3).

Table 2 IPA for customer satisfaction

Item Customer satisfaction

High importantance High performance

Q1 appearance 0.316 6.781

Q2 color 0.326 6.891

Q3 cushion comfort 0.253 6.635

Q4 brake system 0.232 6.760

Q5 shift system 0.270 6.745

Q6 wheel set and transmission 0.188 6.828

Q7 weight 0.167 6.557

Q8 accessories 0.152 6.604



32

Table 3 IPA for customer dissatisfaction

Item Customer dissatisfaction

Low important Low performance

Q1 appearance -0.288 6.158

Q2 color -0.178 6.347

Q3 cushion comfort -0.103 6.116

Q4 brake system -0.286 6.200

Q5 shift system -0.296 6.126

Q6 wheel set and transmission -0.381 6.190

Q7 weight -0.217 5.937

Q8 accessories -0.280 6.105

Moreover, Fig. 1 shows the dual IPA matrix. Q3 was a must-be quality attribute, which

unidirectionally influenced customer dissatisfaction, located in ‘Concentrate Here’

quadrant of the IPA for satisfaction, implying low performance on high importance

attributes. Thus, the enhancement in satisfaction caused by improvement in Q3 was

futile. Similarly, although Q8 was an attractive attribute and unidirectionally influenced

customer satisfaction, its insufficient performance did not lead to customer

dissatisfaction. Thus, because Q3 and Q8 did not result in the expected performance,

improvements in these areas were considered unnecessary. One-dimensional

attributes, namely Q1, Q2, Q4, Q5, and Q6, influenced customer satisfaction as well

as customer dissatisfaction. In the IPA, Q1, Q2, Q4, and Q5 indicated satisfaction;

subsequently, Q1, Q4, and Q6 indicated dissatisfaction. Thus, Q1, Q2, Q4, and Q5

should be improved to enhance satisfaction; and Q1, Q4, and Q6 should be

maintained to prevent dissatisfaction.

5.8

5.9

6.0

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9

7.0

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4

Q4Q1

Q2

Q5

Q3

Q6

Q7Q8

Q1

Q2

Q3

Q4

Q5

Q6

Q7Q8

Perf

orm

ance

Importance Dissatisfaction Satisfaction



33

Fig. 1 A Dual IPA Matrix

Customer Satisfaction Values

High and low importance was used to measure outcomes of gain and loss with a

value function; these values facilitate evaluating customer satisfaction probabilities.

Table 4 details the value derived from the importance values. The value function is

defined as follows:

0)(

0)(

xifx

xifxxv

(1)

where α and β are between 0 and 1; λ > 1; x > 0 indicates gain; and x < 0 indicates

loss. For our results, α = β and λ (Tversky & Kahneman, 1992).

Table 4 Customer satisfaction values

Quality elements Value

Gain Loss

Q1 appearance 0.363 -0.752

Q2 color 0.373 -0.493

Q3 cushion -0.304

Q4 brake system 0.276 -0.748

Q5 shift system 0.316 -0.771

Q6 wheel set and transmission 0.230 -0.962

Q7 weight

Q8 accessory 0.191

DISCUSSION

In the preceding case study, the quality attributes of a bicycle were categorised with

the Kano model. Table 1 shows the Kano quality attributes as well as the impact

levels. The importance of customer satisfaction indicates the impact of a quality

element on customer satisfaction. The attributes arranged in descending order of

customer satisfaction importance were Q2, Q1, Q5, Q4, Q6, and Q8, whereas those

arranged in descending order of customer dissatisfaction were Q6, Q5, Q1, Q4, Q2,

and Q3. Compared with the improvement in Q8 (accessories) performance,

improvement in Q2 (colour) performance significantly increased customer satisfaction.

Therefore, the highest priority must be given to improvement in Q2, rather than Q8

(an attractive quality) in order to improve customer satisfaction. By contrast, Q3, a



34

must-be quality, need not be improved first to eliminate customer dissatisfaction;

however, Q6, with the highest importance of customer dissatisfaction, should the first

instead.

Dual IPA considers the asymmetric relationship between attribute-level performance

and customer satisfaction. A more detailed understanding of the relationship between

attribute-level performance and satisfaction can be observed Tables 2 and 3 and Fig.

1, which highlight the differences between satisfaction and dissatisfaction. As shown

in Fig. 1, attributes located in ‘Concentrate Here’ could not be improved first. If

attractive attributes are in the quadrant of dissatisfaction or must-be attributes are in

quadrant of satisfaction, performance may not impact satisfaction.

Furthermore, Table 4 presents the outcomes of gain and loss. Thaler (1985)

suggested that the outcomes of the prospect theory could be segregated or

integrated to maximise value. According to segregation of multiple gains,

improvement in Q2 and Q5 (0.689) was greater than that in Q1 and Q6 (0.593).

However, because of the integration of multiple losses, Q2 and Q5 demonstrated

lower performance than did Q1 and Q6 (0.474 vs. 0.669). When customers’

requirements are not met, Q1 and Q6 should be improved first; moreover, to further

enhance customer satisfaction, Q2 and Q5 should be improved before Q1 and Q6.

The classification criteria of the Kano model are logical and quantitative through

regression analysis; furthermore, IPA considers the asymmetrical and nonlinear

effects between quality attributes and customer satisfaction. Understanding the

asymmetrical and nonlinear relationship aids in accurately analysing the priority

ranking for improvement. The Kano model and IPA complement each other well for

detailed examination of attributes. To enhance customer satisfaction, prospect theory

provides managers with clear and effective guidance for decision-making.

CONCLUSIONS

In this study, a novel methodology was developed by incorporating the quantitative

Kano model and IPA assessing customer satisfaction. The results of the attribute

classification were more accurate than those of the conventional Kano model and IPA,

which is in agreement with previous studies (Kuo, Chen, & Deng, 2012; Yin, Cao,

Huang, & Cao, 2016). However, no study has explored behavioural economics in this

context. Employing the value function of prospect theory provides greater utility for

prioritising attributes for performance improvement. Improving attractive quality

attributes might not necessarily result in the largest increase in customer satisfaction,

and must-be quality attributes do not strongly influence dissatisfaction. Furthermore,



35

attributes in ‘Concentrate Here’ need not be improved. However, these findings

contradict the reasoning of the Kano model and IPA. Managers should

comprehensively consider whether attributes in ‘Concentrate Here’ require treatment.

Through consideration of behavioural economics, the results indicated that employing

value function aided in determining how decision-makers should prioritise resource

allocation. This study incorporated the Kano–IPA model and prospect theory, without

any limitations; nevertheless, the potential application of this combination requires

further explanation and validation.

This study provides a quality attribute categorisation methodology for precisely

fulfilling customer needs and helping managers make strategic decisions to satisfy

customers. Quality attributes can change over time; in other words, a successful

attribute follows a lifecycle from indifferent quality to attractive quality to

one-dimensional quality, finally to must-be quality (Kano, 2001). Forecasting the

changes in quality attributes is crucial for providing a quality that fulfils customer

needs just in time. Thus, a more precise dynamic two-dimensional quality model will

be constructed to then increase the advantage of a product or service in highly

competitive global market.



36

REFERENCES

Atlason, R. S., Oddsson, G. V., & Unnthorsson, R. (2014). Geothermal power plant maintenance:

Evaluating maintenance system needs using quantitative kano analysis. Energies, 7(7), 4169-4184.

Dominici, G., Roblek, V., Abbate, T., & Tani, M. (2016). Click and drive: Consumer attitude to product

development: Towards future transformations of the driving experience. Business process

management journal, 22(2), 420-434.

Herzberg, F., Mausner, B., & Snyderman, B. B. (1959). The motivation to work. New York: Wiley.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.

Econometrica: Journal of the econometric society, 263-291.

Kano, N, Seraku, N, Takahashi, F & Tsuji, S (1984). Attractive quality and must-be quality, The Journal

of the Japanese Society for Quality Control, 14(2), 39-48.

Kano, N. (2000). Life cycle of quality and attractive quality creation. Proceedings of Quality Excellence

in New Millennium – The 14th Asia Quality Symposium 2000 Taipei, pp.7-11.

Kano, N. (2001). Life cycle and creation of attractive quality. In 4th QMOD Conference, Linköping,

Sweden (pp. 18-36).

Kuo, Y. F., Chen, J. Y., & Deng, W. J. (2012). IPA–Kano model: A new tool for categorising and

diagnosing service quality attributes. Total Quality Management & Business Excellence, 23(7-8),

731-748.

Martilla, J. A., & James, J. C. (1977). Importance-performance analysis. The journal of marketing,

77-79.

Matzler, K., & Renzl, B. (2007). Assessing asymmetric effects in the formation of employee

satisfaction. Tourism Management, 28(4), 1093-1103.

Matzler, K., Bailom, F., Hinterhuber, H. H., Renzl, B., & Pichler, J. (2004). The asymmetric relationship

between attribute-level performance and overall customer satisfaction: a reconsideration of the

importance–performance analysis. Industrial Marketing Management, 33(4), 271-277.

Meng, Q., Jiang, X., & Bian, L. (2015). A Decision-Making Method for Improving Logistics Services

Quality by Integrating Fuzzy Kano Model with Importance-Performance Analysis. Journal of Service

Science and Management, 8(03), 322.

Mittal, V., Ross, W. T., & Baldasare, P. M. (1998). The asymmetric impact of negative and positive

attribute-level performance on overall satisfaction and repurchase intentions. Journal of marketing, 62,

33-47.



37

Riviere, P., Monrozier, R., Rogeaux, M., Pages, J., & Saporta, G. (2006). Adaptive preference target:

Contribution of Kano’s model of satisfaction for an optimized preference analysis using a sequential

consumer test. Food Quality and Preference, 17(7), 572-581.

Shahin, A., Pourhamidi, M., Antony, J., & Hyun Park, S. (2013). Typology of Kano models: a critical

review of literature and proposition of a revised model. International Journal of Quality & Reliability

Management, 30(3), 341-358.

Thaler, R. (1985). Mental accounting and consumer choice. Marketing science, 4(3), 199-214.

Ting, S. C., & Chen, C. N. (2002). The asymmetrical and non-linear effects of store quality attributes on

customer satisfaction. Total Quality Management, 13(4), 547-569.

Tontini, G., Søilen, K. S., & Silveira, A. (2013). How do interactions of Kano model attributes affect

customer satisfaction? An analysis based on psychological foundations. Total Quality Management &

Business Excellence, 24(11-12), 1253-1271.

Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of

uncertainty. Journal of Risk and uncertainty, 5(4), 297-323.

Tversky, A., & Kahnemann, D. (1981). The framing of decision and rational choice. Science, 211,

453-458.

Xu, Q., Jiao, R. J., Yang, X., Helander, M., Khalid, H. M., & Opperud, A. (2009). An analytical Kano

model for customer need analysis. Design Studies, 30(1), 87-110.

Yin, J., Cao, X. J., Huang, X., & Cao, X. (2016). Applying the IPA–Kano model to examine

environmental correlates of residential satisfaction: A case study of Xi'an. Habitat International, 53,

461-472.



38

Documents

INCORPORATING PROSPECT THEORY INTO DECISION-MAKING …ijeronline.com/documents/volumes/2016/Nov - Dec 16... · INCORPORATING PROSPECT THEORY INTO DECISION-MAKING FOR CUSTOMER SATISFACTION