Reference Scales of Service Quality and Satisfaction Judgments

Reference Scales of Service Quality and Satisfaction Judgments: A Reconsideration and Research Agenda Stephen L. Vargo University of Maryland

Outline of Presentation • Implied Reference Scales--Models and Issues • Partial results of three Studies of Reference Scale Organization • Research Implications and Directions

The Research Domain • Reference Scale = The mental “ruler” used in making judgments about marketing phenomena (e. g. , service encounters) • Satisfaction • Service-quality • Components: • Standard: A point on the reference scale that the consumer uses in the comparison of external stimuli • Categories (zones, latitudes): Similarly valenced ranges (latitudes) on a reference scale • associated with common (or similar) evaluative judgments (e. g. , good or bad).

Implied Reference Scales a. Disconfirmation model (single standard) Positive Disconfirmation Negative Disconfirmation (-) (+) Implied Positive Latitude Implied Negative Latitude Expected b. Zone of tolerance model (multiple standard, bounded range) Negative Disconfirmation Latitude of acceptance (Positive Disconfirmation ? ) (-) (+) Implied Negative Zone of Tolerance Minimum Tolerable Standard Desired Standard

Some Problems with Disconfirmation and Zone Models • What is appropriate single standard • e. g. expected, desired, deserved, or adequate performance (Bolton and Drew 1991; Boulding et al. 1993; Spreng and Mackoy 1996) • Nature of Standard (and therefore comparison) • e. g. , “vector attribute” or “ideal point” (Teas 1993) • Expectations and perceptions not independent • “we see what we expect to see” (Pieters, Koelemeijer, and Roest 1996)

Implied Reference Scales (2) Latitude model of social judgment theory (multiple standard, anchor-based) Latitude of objectionability Latitude of acceptance (-) Latitude of objectionability (+) Most Objectionable Most acceptable Latitude of Noncommitment

Reference Scale Types and Issues • Major reference scale models • Single-standard comparison models—e. g. , disconfirmation model (Oliver 1980 ) • Boundary-driven, zone models—e. g. , zone of tolerance model (Parasuraman et al. 1994 ; Woodruff, et al 1983 ) • Anchor-based, latitude models—e. g. , latitude of acceptance (Social judgment theory) (Anderson 1973; Miller 1977 ) • Issues • Nature of standards--boundaries vs. anchors • Related to vector attributes vs. “ideal points” • Relative role of alternative standards • Predicted expectations, normative expectations (should, deserve) desire, minimum tolerable • Dynamics of reference scales under varying conditions • e. g. , prior positive or negative evaluation • e. g. , changes under hi and low involvement conditions

Study 1: Summary of Hypotheses • Consumers differentiate among standards—i. e. , standards play different roles • Standards are not equivalent to latitude boundaries. • Standards consistently associated specific latitudes. • Existence of “hyperservice”—positively rated attribute dimension is evaluated “unacceptable” • Positive and negative behavioral intentions associated with positive and negative latitudes, respectively. • No behavioral intensions associated with neutral latitude.

“Own Categories” Sample Statements: Friendliness • The waitperson tells you that you were wonderful customers (11). • The waitperson writes a personal note of thanks on the check (11). • The waitperson says: "Let me know when you have made up your mind" (6). • The waitperson touches you when talking to you (6). • The waitperson asks a lot of personal questions (6). • The waitperson points out the least expensive items on the menu (6). • The waitperson comments that your clothes are out of fashion (1). • The waitperson swears at you (1). Notes: Number (1 -11) equals median placement in pre-test; Approximately 50 total statements. Also used serving response-time—separate instrument

Category Cards: Waitperson Friendliness a. Waitperson Friendliness 1 Extremely Friendly 2 3 4 5 6 7 8 9 10 11 Extremely Unfriendly b. Serving Response-time 1 Extremely Slow 2 3 4 5 6 7 8 11 Extremely Fast

“Own Categories” Procedures (1) • Respondents given scenario • Lunch with acquaintance in new, unknown family restaurant • Appearance, prices, quantity and quality of food were defined as “as expected” • Asked to sort statements about a dimension (e. g. , friendliness, serving time ) according to similarity • Asked to identify: • Stack which represents most acceptable service level • All other stacks that represent acceptable service levels. • Stack that represents most undesirable service level • All other stacks that represent unacceptable service levels.

“Own Categories” Procedures (2) • Asked to identify • Stacks that represent the service levels they would expect, desire, deserve, find minimally tolerable • Stacks they would associate with various behavioral intentions: • • Positive and negative word-of-mouth (tell friends) Leave Complain Repeat patronage

Average Reference Scale Organization a. Waitperson Friendliness Noncommitment Acceptance Objectionability (Hyperservice) (-) (+) Most Objectionable Minimum Tolerable Deserved Expected Most Acceptable Desired b. Serving Response Time Noncommitment Acceptance Objectionability (Hyperservice) (-) (+) Most Objectionable Minimum Tolerable Deserved Expected Most Acceptable Desired

Research Findings: Placement of Standards and Behavioral Intentions • “Expected” (Expect/Deserve) and “Desired” (Desire/Most acceptable) service standards associated with LA • Do not serve as boundaries • Minimum tolerable associated with LNC • not lower bound of LA • Strong evidence of “hyperservice” • Negative behavioral intentions associated with LO • Positive behavioral intentions associated with LA

Study 2: Focus Group • Purpose • Elaboration and enrichment of quantitative study • Exploratory • Approach • Think about and discuss good and bad restaurant experiences • Think about and discuss meaning of standards (e. g. , desired, ideal, expected, etc. ) • Place standards on two “rulers”—acceptable/unacceptable & friendly/unfriendly—and discuss • Watch video employing subset of statements from card sort • List and number behaviors • Position numbers on rulers (friendliness and acceptability/unacceptability) • Discuss likely responses to behaviors

Study 2: Focus Group: Outcomes • Fairly consistent ordering of standards • Some tendency to equate expected/deserved & desired/ideal • Tendency to group—stack (or “would have stacked if I knew I could”)--behaviors • A lot of support (verbal & on scale) for “hyperservice” • e. g. , flirting is extremely friendly, but unacceptable • Some evidence of different RS organ. Under different conditions • Importance (involvement) • Previous evaluation/relationship

a. Attribute Dimension Typical Mapping of Reference Scales Minimum Tolerable 16 17 Extremely Unfriendly 6 4 11 12 7 9 Expected & Deserved 2 13 1 15 8 Ideal Desired 10 5 Extremely 14 3 Friendly a. Evaluative Dimension 15 14 7 17 Extremely Unacceptable # 6 16 4 13 12 9 11 2 8 = Respondent-observed restaurant behavior from video Minimum Tolerable 1 Expected & Deserved Desired Ideal 10 5 3 #15 = waitperson gave phone # #14 = waitperson sat down when friend left table Extremely Acceptable

Study 3: Experiment • Purpose • Investigate impact of relationship/branding on reference scales • Differences in reference scales for restaurant with prior positive brand evaluation vs. new restaurant • Hypotheses--With prior, positive brand relationship: • • Decreased latitude of objectionability Increased latitude of non-commitment No change in latitude of acceptance Shift of positive BI from LA only to LA and LNC • Method • “Electronic” Card sort • “Branded” scenario assigned to half the respondents • Restaurant is new but recognize brand as part of favorite chain • Outcomes • Hypotheses generally supported (significance and trends) • Both LA and LO decreased • BI associate with both LA and LNC for Branded condition

Study 3: Comparison of Unbranded and Branded Reference Scales Tests of Mean Differences Among Latitude Measures Measure Mean (Branded) Mean (unbrand ed) Difference t-value p Sig Latitude of rejection (size) 3. 00 4. 06 -1. 06 -2. 45 0. 017 Y Latitude of acceptance (size) 3. 07 4. 11 -1. 04 -2. 25 0. 027 Y Latitude of non-commit. (size) 4. 32 2. 17 2. 15 2. 73 0. 008 Y Density of lat. of accept 12. 68 14. 86 -2. 17 -1. 15 0. 252 N Density of lat. of object 16. 44 22. 49 -6. 05 -2. 72 0. 008 Y Density of lat. of non-commit 25. 88 17. 66 8. 22 2. 38 0. 020 Y Minimum tolerable position 5. 00 0. 00 1. 00 N

Study 3: Comparison of Behavioral Intentions Frequency Distributions Of Behavioral Intentions By Latitude Variable Latitude Branded Freq RETURN Acceptability Unbranded Percent Freq Χ 2 df p<. 05 Percent 63 62. 38 71 78. 89 Objectionability 4 3. 96 6 6. 67 Noncommitment 34 33. 66 13 14. 44 9. 96 2 yes

Extensions and Research Agenda • Synthesis • Simultaneous influence of multiple standards • Sorting out the “latitudes” and “zones” in marketing literature • Adds depth to Social Judgment model • Role of minimum tolerable—”adaptation level” • Explanation of Satisfiers and Dissatisfiers • Distributions of expectations, desires • “Six Sigma” • Other evaluative reference scales—price • International Issues • Cross cultural differences in reference scales • Reference scale as index of acculturation

Latitude Relationships Latitude of Non-Acceptability Latitude of Objectionability (intolerance) Latitude of Tolerance Latitude of L of N Non- commitment Acceptance (indifference) L of O (-) Composite Latitudes Primary Latitudes (+) Hyperactivity Latitude Anchors

Latitude Profiles of Service Dimensions

Evaluative Distributions Distribution of objectionability Distribution of Desirability Distribution of Expectations (-) Distribution of objectionability (+) Most Objectionable Expected Desired Latitude of Noncommitment Partially based on Rust, Roland T. et al, (1999) “What You Don’t Know About Customer-Perceived Quality: The Role of Customer Expectation Distributions, Marketing Science” 18 (1), 77 -92.

Implications of Distributions Distribution of Expectations Distribution of Desirability (+) D Desire “Relationship” ↑as D ↓ or, more precisely, Expect Ideal Relationship = Distribution of expectations within distributions of desirability Six Sigma = 99. 9998 % of performance within Latitude of Acceptability

Other Directions • Evaluative reference scales in price research • “altitude of price acceptance” • Cultural Issues • Cultural differences in reference scales • Satisfiers and dissatisfiers as bases for local vs. global • Reference scales as indices of acculturation

Management Implications • Managing the service-encounter • Not sufficient to know what consumer wants • Must know what consumer finds objectionable • Too much service (hyperservice”) may be more harmful than too little • Managing the evaluation process • Competitive advantage through expectations management • Can not manage positive latitudes only • May be more important to manage LO and LNC

Implied Management Strategies