5. QUANTITATIVE ANALYSIS OF THE CASE UNIT
5.2 The Factors of the Product Platform Development
In order to find out the structure behind the statements, i.e. the factors of product platform development, factor analysis was conducted. From the factor analysis, the statements with large amount of missing data (>20%) and high skewness or kurtosis values (<-1 or <1) were excluded. The extraction method used in the factor analysis was the Principal Component Analysis, and the rotation method was the Varimax with Kaiser Normalization. The factor analysis was first conducted simultaneously to all the statements chosen to the further investigation. The results showed enough correlations between the statements (each statement had correlation over 0.3 at least with one another statement). The Measure of Sampling Adequacy (MSA) values were at acceptable level (well over 0.5), but the communalities of the statements as well as the loadings in the factor analysis were too low for some statements. Hence, the factor analysis was conducted several times excluding the statements, which did not fulfill the criteria, from the analysis one by one until all the loadings and communalities in addition to the MSAs (Table 12) and correlations (Appendix 3) were on acceptable level. The analysis resulted in six factors comprised of 19 statements (hence 6 statements were excluded from the analysis) (Table 13). The factors were normally enough distributed, i.e. the skewness and kurtosis values were on acceptable level (Table 14). From the possible 134 cases, 68 were included into the analysis: other cases included one or more missing value(s). The resulting model fit was over middling, the Kaiser-Meyer-Olkin Measure of Sampling Adequacy was 0.75 and the Bartlett's Test of Sphericity Χ2(171) was 572.35 (p<0.001).
Factor analysis
Communalities and MSA-values of the statements included into the Factor analysis
Communality Measure of Sampling Adequacy s_2.3 SWP/R&D strategies /road maps meeting your strategic /
business needs 0.589 0.851
s_9.1 The level SWP takes customers'/your Product Line's needs into
account in product development 0.612 0.842
s_5.1 Functioning of prioritization/decision making 0.585 0.838 s_1.3 SWP/R&D people understanding your business operations 0.583 0.688 s_2.1 SWP/R&D ability to develop clear long-term direction of where
SWP is going (i.e. road maps) 0.708 0.796
s_6.1 Receiving needed information of projects in SWP (visibility to
projects) 0.708 0.737
s_6.2 Quality of SWP project documents / meeting minutes 0.716 0.697
s_10.3 Quality of documentation 0.513 0.856
s_6.4 Delivery completeness 0.623 0.736
s_4.1 SWP informing Product Lines of SWP/R&D contact persons 0.666 0.641 s_4.3 SWP/R&D people participating in your Product Line's meetings
/ reviews when asked 0.614 0.751
s_4.2 Getting responses/help from SWP/R&D 0.603 0.784
s_3.2 Innovativeness of new features from SWP 0.748 0.557
s_3.1 Innovativeness in Product platform design and development
(architecture) 0.801 0.638
s_10.1 Reliability and quality of software 0.679 0.778
s_8.2 Software corrections 0.7 0.81
s_10.2 Reliability and quality of hardware 0.647 0.588
s_5.3 Promised schedules of product platform products (Lead time) 0.747 0.803 s_9.3 DX200 product platform meets your needs/end customers' needs 0.6 0.807
Table 12. Statements included into the factor analysis.
Factor analysis Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.75
Bartlett's Test of Sphericity, Χ2 (171)= 572.35, p=
0.000
Strategic and business fit of product platform Project communication and deliverables Cooperation with product platform development Innovativeness of product platform architecture and features Reliability and quality of product platform Promised schedules and final product platform meeting the needs Eigenvalue – initial Rotated sums of squared loadings
% of Variance s_2.3 SWP/R&D strategies /road maps meeting your
strategic / business needs 0.737 0.010 -0.078 0.171 0.087 0.053
s_9.1 The level SWP takes customers'/your Product
Line's needs into account in product development 0.727 0.094 0.182 -0.001 0.138 0.146 s_5.1 Functioning of prioritization/decision making 0.678 0.197 0.235 0.125 -0.087 0.091 s_1.3 SWP/R&D people understanding your business
operations 0.660 -0.103 0.175 -0.183 0.121 0.242
s_2.1 SWP/R&D ability to develop clear long-term
direction of where SWP is going (i.e. road maps) 0.616 0.430 -0.128 0.345 0.055 -0.075 s_6.1 Receiving needed information of projects in
SWP (visibility to projects) 0.190 0.811 0.044 0.068 -0.051 0.069
s_6.2 Quality of SWP project documents / meeting
minutes -0.080 0.759 0.356 0.033 0.081 -0.006
s_10.3 Quality of documentation 0.113 0.556 0.146 -0.054 0.367 0.177
s_6.4 Delivery completeness 0.022 0.511 -0.105 0.146 0.441 0.367
s_4.1 SWP informing Product Lines of SWP/R&D
contact persons -0.042 0.061 0.765 0.175 -0.077 0.196
s_4.3 SWP/R&D people participating in your Product
Line's meetings / reviews when asked 0.141 0.248 0.728 -0.026 0.003 0.044 s_4.2 Getting responses/help from SWP/R&D 0.206 0.004 0.704 0.063 0.197 -0.149 s_3.2 Innovativeness of new features from SWP 0.051 0.024 0.068 0.854 0.066 0.081 s_3.1 Innovativeness in Product platform design and
development (architecture) 0.188 0.112 0.131 0.852 0.091 0.017
s_10.1 Reliability and quality of software 0.110 0.099 -0.029 0.086 0.787 0.172
s_8.2 Software corrections 0.296 0.339 0.236 0.021 0.621 -0.238
s_10.2 Reliability and quality of hardware -0.091 -0.388 0.078 0.442 0.534 -0.020 s_5.3 Promised schedules of product platform products
(Lead time) 0.303 0.233 0.168 -0.008 0.023 0.756
s_9.3 DX200 product platform meets your needs/end
customers' needs 0.432 -0.031 -0.101 0.292 0.228 0.514
Table 13. Factors.
The factors were straightforward to name. Factor 1 consisted of strategic issues, the business fit of the product platform strategies and taking the product lines’ needs into account, and hence it was named to “Strategic and business fit of product platform”.
The second factor included the project information in different forms, as well as the resulted delivery completeness, and the factor was named to be “Project communication and deliverables”. There was one peculiar statement included into the factor 2: the quality of documentation. The documentation can be thought as being a part of the project’s result, i.e. the release, and hence it can be seen as an integral part of the factor 2 “Project communication and deliverables”. The factor 3 included the contact person information as well as receiving help and participating into the meetings, and therefore the factor name was chosen to be “Cooperation with product platform development”.
The factor 4 included both the questions about the product platform innovativeness, thus the name “Innovativeness of product platform architecture and features” was given to the factor. Factor 5, named to be “Reliability and quality of product platform ”, consisted of the reliability and quality of software and hardware, as well as the software corrections. Factor 6 comprised of two statements, namely the statements about the promised schedules and the product platform meeting the needs. The factor name,
“Promised schedules and final product platform meeting the needs”, included the both aspects of the factor. It is strange though, that the product platform meeting the needs should be combined with the schedules needed to produce the product platform – but in the eyes of the product line, getting the product in right schedule might be a part of fulfilling their needs.
The sample of 134 cases was not large enough to validate the factor analysis by splitting the database into two (if the absolute minimum for factor analysis is 5 cases for each variable in the analysis, meaning 95 cases). Hence, a weaker validation method was chosen. From the database about 80% of the cases were randomly selected – depending on the variable, the number of cases included into the analysis ranged from 81 to 102 – hence the sample size was about large enough to fulfill the minimum required. The single statement MSAs were on acceptable level, and there were enough correlations
between the statements. For the entire model, the KMO was 0.74 compared to the original one’s 0.75, and Bartlett's Test of Sphericity Χ2(171) was 520 (p<0.001). The communalities were well over 0.5, and the total cumulative variance explained was 68.8 (compared 65.5). Only one variable loaded different factor than in the original analysis, namely the factor “8.2 Software corrections” loaded the factor 2 “Project communication and deliverables” by 0.534, but it loaded also significantly the factor 5
“Reliability and quality of product platform ” (0.468), which was loaded more in the original analysis. The variable can be justified to be in either of the factors. Hence, with only small changes, the validation strengthens the results of the original analysis, even though it is not a fully robust validation. Because the factor analysis did not include all the statements, the results of the factor analysis were not expected to be perfect, either.
The bolded factor loadings from the Table 13 were used to create summated scales to be able to compare the factors with one another. The descriptives of the factors and the summated scales are presented in Table 14. The Cronbach’s Alphavalues calculated for the summated scales were not all on the acceptable level (i.e. over 0.6), and hence it was decided to use mainly the factors in the further analysis. The low Cronbach’s Aplha value of factor 6, “Promised schedules and final product platform meeting the needs”, indicates of low reliability of the factor. The summated scales will be used only to analyze the differences between the factors with the t-test. The normality measures of the factors as well as the summated scales were on acceptable level. In addition, the Pearson correlation coefficients between the summated scales and factor scores were high: they ranged from 0.78 to 0.95 (p<0.001). Hence, the summated scales can be used to compare the scores with one another.
Factor / Summated Scale
N Mean Std. Deviation Skewness Std. Error Kurtosis Std. Error Cronbach's Alpha
Factor 1: Strategic and business fit of product platform
68 -0.02 0.99 -0.77 0.29 0.73 0.57 Factor 2: Project communication and
deliverables
68 -0.13 1.00 0.00 0.29 -0.36 0.57 Factor 3: Cooperation with product platform
development
68 0.15 0.97 0.25 0.29 -0.33 0.57 Factor 4: Innovativeness of product platform
architecture and features
68 0.08 0.84 -0.19 0.29 -0.90 0.57 Factor 5: Reliability and quality of product
platform
68 -0.09 1.06 -0.31 0.29 -0.22 0.57 Factor 6: Promised schedules and final product
platform meeting the needs
68 0.02 0.92 -0.10 0.29 -0.27 0.57 sum 1: Strategic and business fit of product
platform
68 2.94 0.58 -0.74 0.29 0.41 0.57 0.77 sum 2: Project communication and
deliverables
68 3.07 0.62 0.12 0.29 -0.43 0.57 0.73 sum 3: Cooperation with product platform
development
68 3.65 0.70 -0.16 0.29 -0.47 0.57 0.65 sum 4: Innovativeness of product platform
architecture and features
68 3.28 0.61 -0.44 0.29 -0.65 0.57 0.82 sum 5: Reliability and quality of product
platform
68 3.17 0.63 -0.26 0.29 -0.67 0.57 0.53 sum 6: Promised schedules and final product
platform meeting the needs
68 3.09 0.71 0.03 0.29 -0.72 0.57 0.50
Table 14. Descriptives of the factors and summated scales.