Beyond Multiplication

Abstract

Objective:

This article brings the discussions on incorporating perceived importance across study areas into the study of client satisfaction and cautions the use of multiplicative scores (multiplying satisfaction and importance scores) as a weighting method. An alternative weighting method is provided.

Method:

Analyze data from a client satisfaction survey of 103 clients of a publicly funded elderly case management service unit located in a large U.S. Midwest region city.

Results:

The alternative weighted score correlated higher with all the global client satisfaction measures than the multiplicative score.

Conclusion:

Using multiplicative scores to represent global client satisfaction should be avoided. The proposed alternative weighting method is a reasonable way to incorporate perceived importance to represent global client satisfaction.

Keywords

instrument construction client satisfaction importance weighting multiplicative score

Introduction

In order to overcome issues precluding social workers and other social service providers from utilizing client satisfaction data for service improvement, Hsieh and Essex (2006) proposed a framework for developing client satisfaction measures. This proposed measurement framework can be achieved by a three-step process: First, identify major elements of care or services provided within the specific setting. Second, construct a Likert-type satisfaction rating item for each of the service elements. Third, construct a Likert-type importance rating item for each of the service elements; or better, construct a mechanism to obtain an importance hierarchy for the service elements. Client satisfaction measures developed using the proposed framework can offer data with direct relevance to the service or practice settings and can pinpoint the sources of satisfaction and dissatisfaction. In addition, data from client satisfaction measures developed using the proposed approach can help social workers and other service providers set priorities for service improvement based on clients’ perceived importance of various service elements (Hsieh, 2006; Hsieh & Essex, 2006).

Despite evidence has shown that perceived importance of service elements is critical in considering client satisfaction (e.g., Hsieh, 2009, 2012a, 2012c), there is no clear guideline on how to incorporate perceived importance to represent global or overall client satisfaction. One intuitive way is to include perceived importance into the scoring of global client satisfaction, otherwise known as importance weighting (Hsieh & Essex, 2006). Importance weighting is conceptually intuitive because it seems reasonable that satisfaction (or dissatisfaction) with service elements that are more important would make larger contributions to global client satisfaction. Unfortunately, importance weighting, though simple in principle, is not so simple in practice. Specifically, it is not clear whether satisfaction in a service element that is “not important at all” should carry zero weight in a client’s global satisfaction, nor is it clear how much more weight should be given for each 1 point increase in a 5-point (or 7-point) importance rating scale. That is, even if the concept of importance weighting is to be accepted, there is no clear answer to how weighting should be implemented.

The Present Study

The same concept of importance weighting has been a topic of focus in several areas of study. As Russell and Hubley (2005) pointed out, investigations of importance weighting can be found in research on self-esteem, life satisfaction, and job satisfaction. The purpose of this article is to bring the discussions on importance weighting across study areas into the study of client satisfaction. Specifically, this article cautions the use of multiplicative scores (multiplying satisfaction and importance scores) as a weighting method and provides an alternative weighting method that has conceptual as well as empirical support. Issues discussed in this article offer practical applications to social workers and other service providers to avoid potential problems in incorporating perceived importance of service elements into client satisfaction.

Major Issues Regarding Importance Weighting

Research on importance weighting continues to receive attention, especially in the life satisfaction literature (e.g., Hsieh, 2003, 2004, 2012b, 2012d, 2013; Rojas, 2006; Russell & Hubley, 2005; Russell, Hubley, Palepu, & Zumbo, 2006; Wu, 2008a, 2008b; Wu & Yao, 2006a, 2006b, 2007). Given the similarities in measurement and conceptualization between life satisfaction and client satisfaction (Hsieh, 2006, 2009; Hsieh & Essex, 2006), lessons learned on the topic of importance weighting in life satisfaction should not be ignored in the client satisfaction context. This article extends the discussions of the main controversies surrounding importance weighting in relation to client satisfaction research that were summarized and discussed previously (Hsieh, 2009, 2012a, 2012c) and gives special attention to the issues of using multiplicative scores as a weighting method for importance weighting.

Over a decade ago, Trauer and Mackinnon (2001) argued against importance weighting by pointing out issues related to multiplicative scores. As they described, a multiplicative score is a score obtained by multiplying a satisfaction score by an importance score. It is common that multiplicative scores are used as a method of importance weighting in life satisfaction and quality of life measures (e.g., Dijkers, 2003; Ferrans & Powers, 1985). The main strength of multiplicative scores is that multiplying satisfaction and importance scores is an intuitive and straight-forward approach to capture the potential interperson differences in perceived domain importance. For example, Ferrans and Powers (1985) constructed the Quality of Life Index to use multiplicative scores as a way to reflect satisfaction with various aspects of life incorporating perceived importance of these life aspects. According to Trauer and Mackinnon (2001), however, the use of multiplicative scores is problematic from both conceptual and psychometric perspectives. Trauer and Mackinnon discussed two conceptual issues regarding multiplicative scores. First, satisfaction items or questions included in any measure were selected based on their implicit importance. Therefore, importance weighting became redundant and unnecessary. Second, the meaning of a multiplicative score is difficult to determine, given it can be obtained by different combinations of satisfaction ratings and importance ratings. Trauer and Mackinnon discussed two psychometric issues regarding multiplicative scores. First, the use of multiplicative scores resulted in a decrease in reliability, based on internal consistency. Second, results from multiplying 2 (satisfaction and importance) item scores that are ordinal level in nature could be unsuitable for statistical analysis. Although the conceptual and psychometric issues Trauer and Mackinnon raised on multiplicative scores and importance weighting were based on the life satisfaction and quality of life literature, these issues have direct implications and relevance to importance weighting in the client satisfaction context. These issues are examined and discussed here:

Issue 1: Importance Weighting Is Unnecessary

The claim that the satisfaction items are chosen to be included in a measure of satisfaction because they are inherently important may be of merit. However, arguing importance weighting is unnecessary due to the inherent importance of the satisfaction items is to make the assumption that all satisfaction items are equally important to everyone (Hsieh, 2012b, 2012d, 2013; Rojas, 2006). In the context of client satisfaction, saying that satisfaction items are all important is the equivalent of saying that all service elements included in the client satisfaction measures are important. Arguing that all service elements are important is reasonable. However, saying all service elements are important does not necessarily mean all service elements are equally important. As Hsieh (2009, 2012a, 2012c) demonstrated in a case management setting, not all service elements were deemed equally important by clients. In fact, the claim that all service elements are inherently important does not necessarily lead to the assumption of equal importance of all service elements. There is no conceptual conflict to allow individual differences in perceived importance of a set of service elements that are viewed important. Given that client satisfaction measures are likely developed by researchers, evaluators, and/or service providers, incorporating individual clients’ perceived relative importance of various service elements into client satisfaction measures can be considered a way to obtain data with a client-centered focus. In sum, the claim that all service elements included in client satisfaction measures have built-in importance does not preclude incorporating perceived importance into global client satisfaction.

Issue 2: Unclear Meaning of a Multiplicative Score

A multiplicative score is the product of simply multiplying a satisfaction score by an importance score. The conceptual ambiguity of a multiplicative score can be easily captured by an example. Let’s assume that there is a client satisfaction measure that asks for satisfaction scores and importance scores for two service elements. Person A has a low satisfaction score of 1 and a high importance score of 5 for each of the service elements. To calculate person A’s multiplicative score which is 10, we multiply the satisfaction score by importance score for each service aspect (1 × 5 = 5, and 1 × 5 = 5) and then take the sum of these products (5 + 5 = 10). It will be impossible to distinguish person A from person B who has a high satisfaction score of 5 and a low importance score of 1 for both service elements, since person B’s multiplicative score will be 10 (5 × 1 + 5 × 1 = 10) as well. The same score, therefore, can mean high satisfaction with unimportant service elements or low satisfaction with highly important service elements. Using multiplicative scores to represent global client satisfaction can be problematic since the meaning of these scores is conceptually ambiguous. However, avoiding the use of multiplicative scores is not the same as abandoning importance weighting completely because multiplicative scores are not the only method of importance weighting (e.g., Campbell, Converse, & Rogers, 1976; Hsieh, 2003, 2004; Hsieh & Essex, 2006). Discarding weighting methods that do not result in conceptual ambiguity as if all weighting methods are multiplicative scores would be unreasonable.

Issue 3: Decreased Reliability

Trauer and Mackinnon (2001) argued that the use of multiplicative scores would lead to a decrease in reliability from the perspective of internal consistency. Although, as indicated previously, it is advisable to avoid using multiplicative scores for importance weighting, this claim of decreased reliability deserves further clarification. The measurement framework of client satisfaction proposed by Hsieh and Essex (2006) is based on a formative-indicator model, not a reflective-indicator model (Chin & Newsted, 1999; P. Cohen, Cohen, Teresi, Marchi, & Velez, 1990; Hsieh & Essex, 2006). In a formative-indicator model, indicators are viewed as determining or causing the construct. In a reflective-indicator model, indicators are viewed as determined by the construct (Bollen & Lennox, 1991; Chin & Newsted, 1999; Cohen et al., 1990). More specifically, measure items or indicators in a reflective-indicator model are considered interchangeable, while measure items or indicators in a formative-indicator model do not need to be interchangeable (Bollen & Lennox, 1991; Cohen et al., 1990). As Bollen and Lennox (1991) suggested, the concept of reliability based on the conventional perspective of internal consistency is appropriate only for a reflective-indicator model. Internal consistency as a measure of reliability does not necessarily apply to a formative-indicator model, given that measure items or indicators in a formative-indicator model do not necessarily share a common theme (Bollen & Lennox, 1991). In other words, the issue of decreased reliability discussed by Trauer and Mackinnon(2001) does not necessarily apply to the client satisfaction measurement framework proposed by Hsieh and Essex.

Issue 4: Measurement Properties of Likert-Type Rating Scale

Another psychometric issue raised by Trauer and Mackinnon (2001) regarding importance weighting has to do with measurement properties. Given that the satisfaction as well as importance scores are typically obtained using Likert-type rating scales, these scores are ordinal in nature. Trauer and Mackinnon argued that scores obtained by multiplying 2 (satisfaction and importance) ordinal-level items would not have the necessary measurement properties for statistical analysis. In fact, the very same issue of measurement properties applies all measures constructed using the Likert-type scales (e.g., Jamieson, 2004). Given that many client satisfaction measures, such as the Client Satisfaction Questionnaire (CSQ-8; Nguyen, Attkisson & Stegner, 1983), the Reid-Gundlach Social Service Satisfaction Scale (R-GSSSS, Reid & Gundlach, 1983), and the proposed measurement framework by Hsieh and Essex (2006), are constructed with Likert-type scale response items, Trauer and Mackinnon’s (2001) questioning of measurement properties may appear reasonable. The controversy regarding measurement properties of Likert-type scales has been in existence for decades (e.g., Carifio & Perla, 2007, 2008; Jamieson, 2004; Norman, 2010). As suggested by Carifio and Perla (2007, 2008) and Norman (2010), concerns about using parametric statistics to analyze data obtained from Likert-type scales are often overexaggerated and unfounded. More specifically, Carifio and Perla and Norman argued that responses combined across multiple items of Likert-type scales should be considered having continuous-level measurement properties, and the use of parametric statistics is appropriate. In addition, most of the common parametric statistics still produce robust results in the event that the assumption of normal distribution is violated (Carifio & Perla, 2007, 2008; Norman, 2010). Unless there are reasons to believe that there are unique issues regarding measurement properties in life satisfaction or client satisfaction studies, the concern of adequacy of Likert-type scale data for statistical analysis is likely a myth (Carifio & Perla, 2007, 2008; Norman, 2010).

In sum, none of the issues discussed by Trauer and Mackinnon (2001) regarding importance weighting is a cause of major concern, except for the conceptual ambiguity of multiplicative scores. The critical lesson from life satisfaction and quality of life studies applicable to client satisfaction studies is to avoid the use of multiplicative scores as a weighting method to incorporate importance into global satisfaction. In particular, importance weighting must be constructed without conceptual ambiguity.

Beyond Multiplication

To overcome the issue of conceptual ambiguity as a result of multiplicative scores, a simple alternative weighting method could be used. Instead of the multiplication of satisfaction and importance scores, weighting is achieved by including the sum of importance scores across all domains as a denominator (see Hsieh, 2003; Hsieh & Essex, 2006). That is, an individual’s global client satisfaction score is:

\sum (S_{i} \times I_{i}) / \sum I_{i},

where S_i is the satisfaction score of service element i and I_i is the importance score of service element i.

Unlike multiplicative scores that may be difficult to interpret, the score obtained by this alternative weighting scheme clearly indicates an individual’s weighted (using importance) satisfactions. The hypothetical example used earlier would illustrate the case. While multiplicative scores fail to differentiate person A from person B, the alternative weighting method clearly shows the difference that person A has a score of 1, (1 × 5 + 1 × 5)/(5 + 5) = 1, and person B has a score of 5, (5 × 1 + 5 × 1)/(1 + 1) = 5. Below, a reanalysis of an earlier study (see Hsieh, 2006, for details) is used as an empirical example to further illustrate the conceptual ambiguity of multiplicative scores and the adequacy of the alternative weighting method proposed.

Method

Data and Participants

A client satisfaction survey was conducted at a publicly funded elderly case management service unit located in a large city in the Midwest region of the United States. Study participants were randomly selected among the clients who were scheduled for the Unit’s follow-up or reassessment visits (so the Unit’s staff could obtain face-to-face written consent to be contacted for research) during the months of January through July 2005. Clients who could not speak English were excluded from the study. Clients who could not speak English were excluded from the study, given that the client satisfaction measure was constructed in English and the interviews were conducted in English. Clients who scored lower than 21 on the Mini-Mental State Examination (MMSE; Folstein, Folstein, & McHugh, 1975) were also excluded to avoid any potential problems due to cognitive impairment.

Procedures

Upon the receipt of client’s consent to be contacted, a trained research assistant who was a student of master of social work program from a local university set up interview appointments. Written informed consents for the research were obtained, and face-to-face interviews were conducted by the research assistant at the homes of the participants. Fourteen of the 141 clients who gave consent to be contacted could not be reached by the research assistant (after at least five attempts). Fifteen of the remaining 127 clients refused to participate in the study. A structured face-to-face survey questionnaire was used for the interviews conducted at the homes of the participants. Interviews lasted, on average, 20 min, and participants received US$10 for each interview. The study was approved by the University of Illinois at Chicago’s institutional review board. A total of 112 interviews were completed at the end of September 2005. After excluding those with missing data, the sample size for the current study was 103. These 103 respondents had complete data on all the variables (including demographic variables) used in the analysis in this study. Most of the study participants were female (81%) and African American (93%). The mean age of the study participants was 75.77 (SD = 7), ranging from 62 to 93. The mean years of schooling completed were 9.77 (SD = 3), ranging from 2 to 16. Most of them were retired (96%) and had an annual household income below US$15,000 (94%). Table 1 shows the major demographic characteristics of the respondents.

Table 1.

Major Demographic Characteristics of Study Participants.

Variables	n (%) or M (SD)	Range
Gender
Female	83 (80.6)
Male	20 (19.4)
Race
African American	96 (93.2)
Other	7 (6.8)
Age	75.77 (7.0)	62–93
Education (years of schooling completed)	9.77 (3.0)	2–16
Marital status
Married	7 (6.8)
Widowed	55 (53.4)
Divorced	23 (22.3)
Separated	12 (11.7)
Single, never been married	6 (5.8)
Annual household income
US0–$4,999	4 (3.9)
US$5K–$9,999	64 (61.5)
US$10K–$14,999	29 (28.2)
US$15K–$19,999	4 (3.9)
US$20K–$24,999	2 (1.9)

Note. N = 103. SD = standard deviation.

Measurement and Analysis

Measures of Service Element Satisfaction and Importance

Participants were asked to rate their satisfaction rating for each of the five major service elements (assessment of needs, plan of care development, case manager’s knowledge regarding available services, case manager’s ability to get services, and the availability of the case manager) with 7-point Likert-type rating items. The statement used for the Likert-type satisfaction rating was “Please use a number from 1 to 7 to indicate your satisfaction where 7 means completely satisfied and 1 means completely dissatisfied. If you are neither completely satisfied nor completely dissatisfied, you would put yourself somewhere from 2 to 6; for example, 4 means neutral, or just as satisfied as dissatisfied.” Participants were asked to rate the importance of each of the five service elements, using the question: “Some people may feel some areas of the case management services are more important than others. What areas of case management services do you consider extremely important or not at all important to you? Please use a number to indicate the importance of the services from 1 through 5, where 5 means ‘extremely important’ and 1 means ‘not at all important’.” The 7-point rating scale has been a choice for measuring satisfaction and the 5-point rating scale has been a choice for measuring importance since the 1970s (e.g., Campbell et al., 1976; Hsieh, 2003, 2006). To increase precision of the importance measure, participants were then asked to compare and rank among the service elements with same rating response options to obtain a rank ordering of service elements. In other words, this produced a hierarchy of service elements from 1 (most important) to 5 (least important) for each respondent. It was possible for service elements of equal importance to be so ranked. Although both rating and ranking data were obtained in the data set, only the ranking data were used in this analysis (given that ranking data could be considered more precise).

Measures of Global Client Satisfaction

Three measures of global client satisfaction were used. One was a single-item question asking the participants to rate their satisfaction with the services all together as completely satisfied (7), somewhat satisfied (6), slightly satisfied (5), neither satisfied nor dissatisfied (4), slightly dissatisfied (3), somewhat dissatisfied (2), or completely dissatisfied (1). The mean rating for this global client satisfaction item was 6.44 (SD = .84), ranging from 2 to 7. The Home Care Satisfaction Measure: Case Management Service (HCSM-CM13; Geron et al., 2000) was also used as a global measure of client satisfaction. The HCSM-CM13 is a global client satisfaction measure designed for case management service settings and has been shown to be valid and reliable (Geron et al., 2000). The reliability coefficient (Cronbach’s α) was .79 for this 13-item measure for the current study sample. In addition, the 8-item Client Satisfaction Questionnaire, CSQ-8 (Nguyen et al., 1983), was used as a measure of global client satisfaction. The CSQ-8 is a popular global client satisfaction measure that has been used in various service settings (e.g., Hsieh, 2006). The reliability coefficient (Cronbach’s α) was .84 for this 8-item measure for the current study sample.

Correlation (Pearson r) analysis was used to compare the associations between global client satisfaction measures and multiplicative scores and global client satisfaction measures and weighted scores based on the alternative weighting method.

Results

Table 2 shows means, standard deviations, and ranges of the multiplicative scores, the weighted scores based on the alternative weighting method, and the three global client satisfaction measures. The mean score of the single-item global client satisfaction rating was 6.44 (SD = .84), ranging from 2 to 7. Given that the highest rating for this single-item satisfaction question was 7, the mean score of 6.44 indicated a high level of satisfaction. The mean score of the CSQ-8 was 28.90 (SD = 3.3), ranging from 19 to 32. Given that the maximum score for CSQ-8 was 32, the mean score of 28.90 indicated a high level of satisfaction. The mean score of the HCSM-CM13 was 54.95 (SD = 6.58), ranging from 38 to 65. Since the highest possible score for HCSM-CM13 was 65, the mean score of 54.95 indicated a high level of satisfaction. The mean score of the multiplicative score was 125.04 (SD = 32.52), ranging from 60 to 175. Since the meaning of a multiplicative score is not clearly defined, it is difficult to make sense what the mean score of 125.04 in relation to satisfaction level. The mean score of the alternative weighted score was 6.32 (SD = 0.61), ranging from 4 to 7. Like the single-item client satisfaction measure, the alternative weighted score had a maximum score 7. The mean score of 6.32 suggested a high level of satisfaction.

Table 2.

Means and Standard Deviations of Satisfaction Rating and Importance Ranking of Service Elements (N = 103).

	M (SD)	Range
Single-item global client satisfaction	6.44 (0.84)	2–7
CSQ-8	28.9 (3.3)	19–32
HCSM-CM13	54.95 (6.58)	38–65
Multiplicative score	125.04 (32.52)	60–175
Alternative weighting	6.32 (0.61)	4–7

Note. N = 103. CSQ-8 = 8-item Client Satisfaction Questionnaire; HCSM-CM13 = Home Care Satisfaction Measure: Case Management Service.

Table 3 shows the correlations between the multiplicative scores, weighted scores, and the three global client satisfaction measures. As shown in Table 3, the correlation between the multiplicative score and the single-item global client satisfaction, CSQ-8, and HCSM-CM13 was .26, .29, and .33, respectively. The correlation between the alternative weighted score and the single-item global client satisfaction, CSQ-8, and HCSM-CM13 was .73, .68, and .58, respectively. That is, correlations between the multiplicative score and the global client satisfaction measures ranged from .26 to .33, indicating a small to medium effect size (J. Cohen, 1988). Correlations between the alternative weighted score and global client satisfaction measures ranged from .58 to .73, indicating a large effect size (J. Cohen, 1988). The alternative weighted score correlated higher with all the global client satisfaction measures than the multiplicative score. Given that the association is stronger, as measured by correlations, between the alternative weighted score and the global client satisfaction measures than the multiplicative score, the alternative weighted score would be a more desirable choice than the multiplicative score in representing global client satisfaction.

Table 3.

Correlations Between Measures of Global Client Satisfaction and Weighted Scores (N = 103).

	Multiplicative Score	Alternative Weighted Score
Single item	.26**	.73***
CSQ-8	.29**	.68***
HCSM-CM13	.33***	.58***

Note. CSQ-8 = 8-item Client Satisfaction Questionnaire; HCSM-CM13 = Home Care Satisfaction Measure: Case Management Service.

**p < .01. ***p < .001.

Discussion and Applications to Social Work

Clients’ perceived importance of various service elements can be useful for service improvement (e.g., Hsieh & Essex, 2006). The way in which perceived importance should be incorporated into client satisfaction is unclear, however. This article brings the discussion of importance weighting in life satisfaction and quality of life studies into client satisfaction studies. Although several reasons have been offered to argue against importance weighting, these reasons do not provide sufficient justifications to abandon importance weighting. The main problem is the use of multiplicative scores as a weighting method. In addition to pointing out that the conceptual meaning of a multiplicative score is uncertain, this study presented an empirical example to assess the adequacy of multiplicative scores. Before discussing findings from the empirical example, major limitations should be addressed. First, data for this study came from a survey of one service setting (case management service unit) with a sample of 103 older adults. Generalizability of these results may, therefore, be somewhat limited. More specifically, study results based on 103 respondents might not be representative of over 4,000 clients of the case management service unit surveyed. The study site that was located in a large metropolitan area might not be representative of all elderly case management service units. In addition, study findings based only on client satisfaction with elderly case management services could not be considered applicable to client satisfaction with other types of social services provided by social workers. Second, the values or scores assigned to importance as well as satisfaction were arbitrary. For example, the values assigned to the 7-point Likert-type scale satisfaction responses in the study followed the conventional range of 1 to 7. It is difficult to justify why the values should take 1 to 7 and not 0 to 6 or some other range of values (e.g., Trauer & Mackinnon, 2001). Despite these limitations, findings are still worth noting and have the following two major implications for client satisfaction research and practice: First, as shown earlier, correlations between the multiplicative score and all global client satisfaction measures were fairly low. That is, there does not appear to be any empirical support to use multiplicative scores to represent global client satisfaction. Therefore, it is advisable to avoid the use of multiplicative scores for importance weighting. Second, to avoid the use of multiplicative scores does not mean that importance weighting should be abandoned. Weighting should be constructed in a fashion without conceptual ambiguity. The alternative weighting method presented in this article is one way to overcome the problem of conceptual ambiguity. As shown in the empirical example, correlations between the weighted scores and various measures of global client satisfaction were all above .5, showing high concurrent validity (Shultz & Whitney, 2004). These results suggest that using weighted scores to represent global client satisfaction is reasonable. These results could also be regarded as empirical support for importance weighting in the client satisfaction context.

Considering the limitations of this current study, findings of this study are by no means conclusive. Given our limited understanding regarding importance weighting in the client satisfaction context, further research on this topic is necessary. Future studies can advance our understanding of the topic of importance weighting in the client satisfaction context by examining client satisfaction in different service settings and/or with other client populations. It is important to note that although it seems reasonable that satisfaction (or dissatisfaction) with service elements that are more important should count more in global client satisfaction, there is no reason to believe that the increase or decrease in importance must follow a linear function like what was presented in this article. It is possible that perceived importance may function in a curvilinear or nonlinear fashion. In quality of life studies, Campbell, Converse, and Rogers (1976), for example, discussed a number of possible approaches to importance weighting, such as “hierarchy of needs,” “threshold,” and “ceiling.” The hierarchy of needs approach suggests certain kinds of needs are more essential than others, and one would then expect that unless these most essential needs are reasonably satisfied, it probably does not matter much in terms of overall satisfaction what happens in the less essential ones. The threshold approach suggests that overall satisfaction depends on the presence of some threshold number of satisfactions. If the number of satisfied domains does not meet this threshold, it is suggested that the person would not feel satisfied as a whole. The ceiling approach suggests there is a top limit or ceiling to the number of satisfactions experienced by an individual, and satisfaction beyond the top limit would not produce increased overall satisfaction (Campbell et al., 1976). Although these approaches were discussed in the context of quality of life studies (Campbell et al., 1976), these approaches may be applicable to the client satisfaction context. Our understanding of the topic of importance weighting can also benefit from further conceptual as well as empirical development of weighting approaches.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Bollen

Lennox

(1991). Conventional wisdom on measurement: A structural equation perspective. Psychological Bulletin, 110, 305–314.

Campbell

Converse

P. E.

Rogers

W. L.

(1976). The quality of American life: Perceptions, evaluations, and satisfactions. New York, NY: Russell Sage.

Carifio

Perla

(2007). Ten common misunderstandings, misconceptions, persistent myths and urban legends about Likert scales and Likert response formats and their antidotes. Journal of Social Sciences, 3, 106–116.

Carifio

Perla

(2008). Resolving the 50 year debate around using and misusing Likert scales. Medical Education, 42, 1150–1152.

Chin

W. W.

Newsted

P. R.

(1999). Structural equation modeling analysis with small samples using partial least squares. In Hole

R. H.

(Ed.), Statistical strategies for small sample research (pp. 307–341). Thousand Oaks, CA: Sage.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Cohen

Teresi

Marchi

Velez

C. N.

(1990). Problems in the measurement of latent variables in structural equations casual models. Applied Psychological Measurement, 14, 183–196.

Dijkers

M. P.

(2003). Individualization in quality of life measurement: Instruments and approaches. Archives of Physical Medicine and Rehabilitation, 84, S3–S14.

Ferrans

C. E.

Powers

M. J.

(1985). Quality of life index: Development and psychometric properties. Advances in Nursing Science, 8, 15–24.

10.

Folstein

M. F.

Folstein

S. E.

McHugh

P. R.

(1975). Mini-mental state: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189–198.

11.

Geron

S. M.

Smith

Tennstedt

Jette

Chassler

Kasten

(2000). The Home Care Satisfaction Measure: A client centered approach to assessing the satisfaction of frail older adults with home care services. Journals of Gerontology: Psychological Sciences and Social Sciences, 55B, S259–S270.

12.

Hsieh

C. M.

(2003). Counting importance: The case of life satisfaction and relative domain importance. Social Indicators Research, 61, 227–240.

13.

Hsieh

C. M.

(2004). Income and financial satisfaction among older adults in the United States. Social Indicators Research, 66, 249–266.

14.

Hsieh

C. M.

(2006). Using client satisfaction to improve case management services for the elderly. Research on Social Work Practice, 16, 605–612.

15.

Hsieh

C. M.

(2009). Importance counts: The role of relative importance of service elements in client satisfaction measures. Journal of Social Service Research, 35, 23–31.

16.

Hsieh

C. M.

(2012a). Incorporating perceived importance of service elements into client satisfaction measures. Research on Social Work Practice, 22, 93–99.

17.

Hsieh

C. M.

(2012b). Importance is not unimportant: The role of importance weighting in QoL measures. Social Indicators Research, 109, 206–278.

18.

Hsieh

C. M.

(2012c). Perceived importance of service elements in client satisfaction measures. Journal of Social Services Research, 38, 529–540.

19.

Hsieh

C. M.

(2012d). Should we give up domain importance weighting in QoL measures? Social Indicators Research, 108, 99–109.

20.

Hsieh

C. M.

(2013). Issues in evaluating importance weighting in quality of life measures. Social Indicators Research, 110, 681–693.

21.

Hsieh

C. M.

Essex

(2006). Measuring client satisfaction among older adults and families. In Berkman

(Ed.), Handbook of social work in health and aging (pp. 1009–1017). New York, NY: Oxford University Press.

22.

Jamieson

(2004). Likert scales: How to (ab)use them. Medical Education, 38, 1212–1218.

23.

Nguyen

T. D.

Attkisson

C. C.

Stegner

B. L.

(1983). Assessment of patient satisfaction: Development and refinement of a service evaluation questionnaire. Evaluation and Program Planning, 16, 109–118.

24.

Norman

(2010). Likert scales, levels of measurement and the “laws” of statistics. Advances in Health Science Education, 15, 625–632.

25.

Reid

P. H.

Gundlach

J. H.

(1983). A scale for the measurement of consumer satisfaction with social services. Journal of Social Service Research, 7, 37–54.

26.

Rojas

(2006). Life satisfaction and satisfaction in domains of life: Is it a simple relationship? Journal of Happiness Studies, 7, 467–497.

27.

Russell

L. B.

Hubley

A. M.

(2005). Importance ratings and weighting: Old concerns and new perspectives. International Journal of Testing, 5, 105–130.

28.

Russell

L. B.

Hubley

A. M.

Palepu

Zumbo

B. D.

(2006). Does weighting capture what’s important? Revisiting subjective importance weighting with a quality of life measure. Social Indicators Research, 75, 146–167.

29.

Shultz

K. S.

Whitney

D. J.

(2004). Measurement theory in action: Case studies and exercises. Thousand Oaks, CA: Sage.

30.

Trauer

Mackinnon

(2001). Why are we weighting? The role of importance ratings in quality of life measurement. Quality of Life Research, 10, 579–585.

31.

C. H.

(2008a). Can we weight satisfaction score with importance ranks across life domains? Social Indicators Research, 86, 468–480.

32.

C. H.

(2008b). Examining the appropriateness of importance weighting on satisfaction score from range-of-affect hypothesis: Hierarchical linear modeling for within-subject data. Social Indicators Research, 86, 101–111.

33.

C. H.

Yao

(2006a). Do we need to weight item satisfaction by item importance? A perspective from Locke’s range-of-affect hypothesis. Social Indicators Research, 79, 485–502.

34.

C. H.

Yao

(2006b). Do we need to weight satisfaction scores with importance ratings in measuring quality of life? Social Indicators Research, 78, 305–326.

35.

C. H.

Yao

(2007). Importance has been considered in satisfaction evaluation: An experimental examination of Locke’s range-of-affect hypothesis. Social Indicators Research, 81, 521–541.