here - ERIM - Erasmus Universiteit Rotterdam

here - ERIM - Erasmus Universiteit Rotterdam here - ERIM - Erasmus Universiteit Rotterdam

23.08.2013 Views

A validation Study of House of Quality key performance indicators The scree plot and eigenvalues provided by SPSS, indicate a composition of four factors. After deciding the number of factors, the communality values were inspected. A communality (hi2) = % of variance in xi explained by the factor model. When a communality value of a variable is relative low (.414. The Varimax rotation procedure was used to select variables into factors. The varimax rotation procedure is an orthogonal method of rotation (axes are maintained at right angles) that minimizes the number of variables with high loading factors on a factor thereby enhancing the interpretability of the factor (Malhotra & Birks, 2004). Component 1 2 3 4 Up-to-date information ,807 Reliability of information ,796 Offering relevant information ,728 Member involvement on development ,481 Interaction between Union and members ,425 ,500 Interaction between members ,674 Involvement third parties ,604 Recognizing needs of the members ,461 Perception of solidarity ,499 Opportunity of physical meetings ,721 Opportunity of virtual meetings ,779 Members in the same position ,639 Good representation of interests ,641 Perception of safety and protection ,509 ,656 Influence of online union ,756 Intelligibility objectives and services ,507 Offering effective support ,646 Reliability and quality of services ,556 Professional presentation ,511 ,632 Good online facilities ,777 Accessibility of information ,608 Custom-made services ,658 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a Rotation converged in 6 iterations. Table 6 - Rotated component matrix Factor analysis, as presented in Table 6 indicates four factors which collectively explained 58,7% of the variance in all items. Based on the exploratory investigations of phase I, the original model ( 58

A validation Study of House of Quality key performance indicators Table 2) consisted five dimensions (factors). The explored dimension “Solidarity” was not recognized as a single dimension. Factor analysis merged the dimensions “Interaction” and “Solidarity” and indicated it as one dimension, that we (still) name “Interaction”. Furthermore the item “recognizing the needs of the members” switched from the dimension “Interaction” to “Information”, whereas the item “reliability and quality of services” was replaced from the dimension “Tangibles” to “Safety”. Due to the high factor loadings and plausible interpretations, we decided to adopt the new structure. The new dimensional structure is presented in Table 7. Scale Reliablity Table 7 - Dimensional structure after Factor Analysis After the factor analysis, item analysis was conducted to test if the dimensions composed by the factor analysis are reliable. Cronbach’s alphas were employed to test the reliability of the dimensions. Item anlysis is conducted after the factor anlysis because Cronbach’s alphas measures how well a set of items measures a unidimensial construct; the scale has no underlying factors. Cronbach’s alphas value varies betwen 0 and 1, where values 0.7 are acceptable. As presented in Table 8, for all dimensions Cronbach’s alphas are >0.7 which implicates that the scales are reliable. 59

A validation Study of House of Quality key performance indicators<br />

The scree plot and eigenvalues provided by SPSS, indicate a composition of four factors. After<br />

deciding the number of factors, the communality values were inspected. A communality (hi2) = % of<br />

variance in xi explained by the factor model. When a communality value of a variable is relative low<br />

(.414.<br />

The Varimax rotation procedure was used to select variables into factors. The varimax rotation<br />

procedure is an orthogonal method of rotation (axes are maintained at right angles) that minimizes<br />

the number of variables with high loading factors on a factor t<strong>here</strong>by enhancing the interpretability<br />

of the factor (Malhotra & Birks, 2004).<br />

Component<br />

1 2 3 4<br />

Up-to-date information ,807<br />

Reliability of information ,796<br />

Offering relevant information ,728<br />

Member involvement on development ,481<br />

Interaction between Union and members ,425 ,500<br />

Interaction between members ,674<br />

Involvement third parties ,604<br />

Recognizing needs of the members ,461<br />

Perception of solidarity ,499<br />

Opportunity of physical meetings ,721<br />

Opportunity of virtual meetings ,779<br />

Members in the same position ,639<br />

Good representation of interests ,641<br />

Perception of safety and protection ,509 ,656<br />

Influence of online union ,756<br />

Intelligibility objectives and services ,507<br />

Offering effective support ,646<br />

Reliability and quality of services ,556<br />

Professional presentation ,511 ,632<br />

Good online facilities ,777<br />

Accessibility of information ,608<br />

Custom-made services ,658<br />

Extraction Method: Principal Component Analysis.<br />

Rotation Method: Varimax with Kaiser Normalization.<br />

a Rotation converged in 6 iterations.<br />

Table 6 - Rotated component matrix<br />

Factor analysis, as presented in Table 6 indicates four factors which collectively explained 58,7% of<br />

the variance in all items. Based on the exploratory investigations of phase I, the original model (<br />

58

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!