Personal income Essay
The notion of using polar adjectives to define the termini of semantic dimensions grew out of research on synesthesia (Karwoski & Odbert, 1938). Those researchers and others that followed found that imagery in synthesis is intimately tied to the language metaphor and that both represent semantic relations (Karwoski, Odbert, & Osgood, 1942; Odbert, Karwoski, & Eckerson, 1942; Stagner & Osgood, 1946). The bipolar adjective technique was tested extensively by Osgood at the University of Illinois.
Three dominant factors emerged from the factor analysis used in those studies, and they appeared consistently in the same order of magnitude (Osgood et al. , 1978). The first factor was clearly identifiable as evaluative by listing scales that have high loadings: good-bad, valuable-worthless, and so forth. The second factor was identified as a potency variable: large-small, strong-weak, and so forth. In general, potency variables have considerable evaluative meaning, but their loadings are generally lower than those on the evaluative scales.
The third factor was identified mainly as an activity variable: fast-slow, active-passive, and so forth. Osgood et al. , 1978, reported a noticeable tendency for activity variables to be associated with positive evaluation. The percentage of total variance and common variance accounted for by these three
Need essay sample on "Personal income"? We will write a custom essay sample specifically for you for only $ 13.90/page
A semantic differential presents a rationale for the measurement of attitude, which is a major dimension of meaning in general (Osgood et al. , 1978). Most authorities agree that attitudes are learned and implicit, and further, are predispositions to respond, particularly in an evaluative sense. It appears safe to suggest, then, that attitude is part of the internal mediational activity that operates between most stimulus and response patterns and that the semantic differential may be used as a generalized attitude scale. Further, all of Osgood’s factor analyses produced a dominant evaluative factor.
It would be reasonable to assume that attitude is a dominant factor in semantic space and that the semantic differential is an index of attitude, as well as a method of attitude assessment (Osgood, Suci, & Tannenbaum, 1970). Customers were asked to indicate their perceptions of the overall quality of the service received on three semantic differential scales items with bipolar adjectives addressing high quality, industry standards and best to worst respectively. Customer satisfaction was measured using three semantic differential scales with bipolar adjectives addressing satisfaction, pleasant, enjoyment and one face scale.
These scales were utilised to capture both the cognitive and the emotional nature of this construct (Hausknecht 89-98). Two aspects of post-purchase behaviour were particularly important; future repurchases and recommendation to others. Three items were used to measure future purchase intentions and two items were used to measure recommendation intentions. Perceived monetary costs were measured using three items and perceived time required in obtaining the service was measured by two items. Perceived value was measured by asking customers to evaluate the overall service in the light of “price paid” and “time spent”.
The questionnaire was pre-tested with twenty consumers prior to the field survey. The purpose of this pre-test was to identify any ambiguous wording, and to discover whether respondents had any difficulties in answering the questions. The appendix contains a list of items used. Data Collection Primary research method was used to collect the data. Data were collected by means of a consumer survey. The fieldworker selected customers prior to entering the chosen restaurant at about ten minute’s interval. The customers were informed about the purpose of the research, and were invited to take part in the study.
If the customers agreed to take part, they were given the questionnaire with a return envelope, and a covering letter requesting them to complete and return the questionnaire within a week. A gift was used to stimulate customer participation. It was estimated that roughly one out of four customers approached agreed to take part in the study. Two hundred and fifty five agreed to participate in the study, but only two hundred and seventeen respondents returned the questionnaire. Among those returned questionnaires, eight were incomplete and were discarded.
The number of responses used for the analysis was 209. The resultant sample consisted of 37. 3 percent males, and 62. 7 percent females. Of these 43. 5 percent were aged 18 – 30, and 44. 7 percent had attained a secondary education. The median personal income was between $10,001 and $15,000. Table 1 presents the demographic profile of the respondents. LISREL 8. 14 was used to assess the hypotheses as it had the ability to estimate the multiple and interrelated relationship whilst accounting for the measurement errors in the estimation process (Hair et al.55-78).
The measurement items identified in the Appendix were specified as the indicators of the respective constructs except post-purchase behaviour. In order to maintain parsimony in the number of indicators in the model, two indices were formed by averaging the three items measuring future purchase intentions and the two items measuring recommendation intentions respectively, and they represented the indicators of the post-purchase behaviour (Hair et al. 55-78).
The confirmatory factor analysis results suggested a reasonably good fit with ? 2 statistic = 155.29 (88 degrees of freedom), GFI = 0. 91, NFI = 0. 94, and RMSR = 0. 057. Although the ? 2 statistic was significant, the GFI, NFI and RMSR values were acceptable (Bentler and Bonett 91-104; Sharma 56-85, Kelloway 55-78). It is documented in the literature that the ? 2 statistic is sensitive to sample size. For large sample sizes, even small differences in the sample covariance matrix are statistically significant, although the differences may not be practically meaningful (Sharma 56-85).
The t-statistics showed that the factor loadings were highly significant (all were greater than 1.96), thus providing evidence of convergent validity (Anderson and Gerbing 125-159). The squared multiple correlations were all met the 0. 50 threshold suggested by Fornell and Larcker (78-96) with the exception of the value item measuring the overall service with respect to the monetary costs (r2 = 0. 49). The Cronbach alpha values for the scales ranged from 0. 71 to 0. 92. As a rule of thumb, the Cronbach alpha value should be at least 0. 70 for a scale to demonstrate internal consistency (Nunnally 34-54).
Table 2 presents the results of the confirmatory factor analysis. The structural model included six constructs, of which three were exogenous and three were endogenous. Table 3 presents the covariance results and correlation coefficients among the constructs. The overall fit of the model was reasonable and the r2 for the structural equations ranged from 0. 59 to 0. 79. The Chi-square statistic was 164. 53 (p=0. 00) with 93 degrees of freedom, GFI = 0. 91, NFI = 0. 93 and RMSR 0. 06. Table 4 shows the parameter estimates of the structural equations.