At the onset of this study, in order to achieve the research objectives and validate the hypothesis, a questionnaire was administered to local Iranian pharmaceutical firms, and the obtained results were subsequently analyzed.
Convenience sampling was employed as the sampling methodology, as the companies under study were identified through the website of the Syndicate of the Owners of Human Pharmaceutical Industries of Iran. Visiting these companies in person, the study included those willing and able to participate by completing the questionnaire. The questionnaire, detailed in
Table 1, encompassed four main categories. The questions within each dimension were derived from reputable studies. The first category consisted of the MO questionnaire (
40), the second category focused on DC (
41), the third category addressed corporate SR (
42), and the fourth category explored IP (
43).
| Dimension | Question No. |
|---|
| 1.1. MO-market intelligence generation | 7 |
| 1.2. MO-market intelligence dissemination | 4 |
| 1.3. MO-responsiveness to market intelligence | 6 |
| 2.1. DC-sensing | 3 |
| 2.2. DC-seizing | 4 |
| 2.3. DC-configuration | 5 |
| 3.1. CSR-economic citizenship | 4 |
| 3.2. CSR-legal citizenship | 6 |
| 3.3. CSR-ethical citizenship | 4 |
| 3.4. CSR-discretionary citizenship | 6 |
| 4.1. Innovative performance-exploratory innovation | 6 |
| 4.2. Innovative performance-exploitative innovation | 5 |
Abbreviations: MO, market orientation; DC, dynamic capabilities; CSR, corporate social responsibility.
To ensure the validity of the questionnaire, two different approaches were employed. The first was face validity, which involved supervisors reviewing the questionnaire's appearance and items and providing feedback, which was incorporated into the questionnaire. The second approach was content validity, in which a group of subject-matter experts evaluated and assessed the questionnaire, making a few changes before approving it.
Regarding the reliability of the research tool, a pilot study was conducted with 30 experts in the research field. The results were analyzed using Cronbach's alpha method to measure internal consistency. The questionnaire demonstrated a Cronbach's alpha coefficient of 0.88, indicating a high level of internal correlation among the questionnaire items. This reliability test was further confirmed in
Table 2, where the presence of thirty experts in the field supported the questionnaire's reliability.
| Variable | Experts No. | Alpha |
|---|
| Value | 30 | 0.88 |
The study population consisted of local Iranian pharmaceutical firms. A total of 387 questionnaires were distributed to the top four members of each company, including managing directors, R&D managers, marketing managers, and sales managers. The selection of companies was based on the Syndicate of the Owners of Human Pharmaceutical Industries of Iran, resulting in 129 companies being included in the study. Out of the distributed questionnaires, 300 were completed by representatives from 100 companies, yielding a response rate of approximately 80%.
The inclusion criteria for participation in the study involved companies being registered with the syndicate. To administer the questionnaire, face-to-face meetings were conducted with each member of the company, where the study questions were explained.
However, to mitigate information bias, which is a significant concern in completing the questionnaires, steps were taken to ensure that the quality of information remains consistent among comparison groups. To prevent evaluation bias, a subcategory of information bias, the questionnaire was divided based on the responder's specialty. Each participant only responded to the questions relevant to their area of expertise. Specifically, managers responsible for marketing, direction, and research and development were assigned separate sections of the questionnaire that addressed specific topics such as direct cost, SR, and IP. Subsequently, the validated questionnaire was distributed to experts from 100 Iranian local pharmaceutical companies. The collected data was analyzed using SPSS 21 and smart PLS 4 software, employing structural equation modeling.
The study process consisted of the following phases:
(1) Designing a questionnaire based on the research method.
(2) Validating the questionnaire.
(3) Collecting information from eligible individuals, including criteria members, through the questionnaire.
(4) Evaluating and analyzing the questionnaire results.
(5) Discussing, concluding, and testing the hypotheses.
Additionally, we conducted a survey involving 30 pharmaceutical experts who possess experience in the pharmaceutical market and have knowledge of launching new medicines into the market.
3.1. Data Analysis
Bartlett's test of sphericity was utilized to examine the hypothesis that “the correlation matrix of observed variables is an identity matrix (
44).” By evaluating the significance level of the chi-squared test, this assessment confirms the absence of correlations among the variables. If the P-value of Bartlett's test is less than 5%, it indicates that the matrix is not identified, suggesting a correlation between the variables and rejection of the null hypothesis. In
Table 3, the test's significance level is presented as 0.001, supporting the rejection of the null hypothesis and indicating a significant correlation between the variables. Hence, with a statistical confidence of 95%, the parametric statistical test can be employed, and the sample size is deemed sufficient for conducting factor analysis based on the Kaiser-Meyer-Olkin (KMO) test result (
45).
| Values |
|---|
| Kaiser-Meyer-Olkin measure of sampling adequacy | 0.882 |
| Bartlett's test of sphericity | |
| Approx. chi-square | 7.542E3 |
| Degrees of freedom | 1689 |
| Significance level | 0.001 |
Based on the KMO test presented in
Table 1, the number of questionnaires appears to be sufficient, and Cronbach's alpha value for each dimension of the questionnaire exceeds 0.7.
Due to the good validity of the questionnaires in testing the hypotheses, we utilized smart PLS software and path analysis for this study. Path analysis is employed to evaluate a system of equations that includes all variables. Path models can encompass multiple dependent variables. In SmartPLS, a path model's parameter can consist of single-parameter constructs. When constructing a score for a variable based on multiple indicators, all indicators are assigned equal weight. In essence, the model focuses on the structural significance of the relationships between the observed variables and the control variables. This type of model is commonly used when one or more parameters mediate the relationship between two other variables (mediation models). Additionally, moderated mediation can also be modeled.