The Mediating Effects of Entrepreneurial Orientation Between Procedural Rationality and Strategic Decision-making Effectiveness

Procedural Rationality and Entrepreneurial Orientation

The current research suggests that procedural rationality, as a way that companies use to increase demand for information processing whilst making strategic decisions, affects the entrepreneurial orientation of these companies. This study further proposes that the value of entrepreneurial orientation depends upon the level of rationality that top management employs within such companies. Prior studies in the existing literature reveal that procedural rationality in strategic decision-making has a positive association with entrepreneurial orientation in companies (Covin et al., 2006). Some researchers think that opportunities become obvious to entrepreneurs who know how to effectively acquire, translate and use information (Anokhin et al., 2011; George, 2020). As a result, entrepreneurial orientation is a strategic orientation that reflects the organizational processes (such as procedural rationality) employed by the firm (Lumpkin & Dess, 1996; Wiklund & Shepherd, 2003). Based on the bounded rationality approach, it is noted that the personal values and cognitive abilities of top management teams substantially affect what rational points of view or decisions should be adopted (Hutzschenreuter et al., 2007). In previous literature, Slater and Narver (1995) showed that systematic rational analysis seemed to enhance the entrepreneurial orientation of companies in two ways. First, procedural rationality in strategic decision-making reduces the chances that companies will quickly move to exploit opportunities without taking advantage of all the benefits related to their current opportunities. Second, procedural rationality in strategic decision-making increases the potential for generative learning that boosts more innovative products and services. Moreover, Shane and Delmar (2004) realize that entrepreneurial organizations that opt for a detailed analysis and planning before starting strategic activities are less likely to fail. This can be explained by the fact that the time span between planning and feedback within these organizations is much shorter than that in conservative organizations. Furthermore, Covin et al. (2006) state that companies that adopt entrepreneurial orientation are more inclined to analyse information about what should be done, so as to successfully make new critical efforts and effectively apply learned lessons in the future. In a similar vein, Hammedi et al. (2011), Wales et al. (2020), Putniņš and Sauka (2020), Meekaewkunchorn et al. (2021) point out that companies that adopt entrepreneurial orientation can achieve higher market performance, if they offer innovative products that follow rational routines.

In light of this, the existing evidence supports the hypothesis that procedural rationality in strategic decision-making may enhance entrepreneurial orientation in firms. Taking these points into account, this study suggests that:

H₁: The greater procedural rationality that is adopted, the better the entrepreneurial orientation within such organizations (see Figure 1).

Entrepreneurial Orientation and the Effectiveness of Strategic Decision-making

Entrepreneurial orientation within organizations has recently attracted the attention of researchers (Dada & Fogg, 2016; Wales et al., 2013; Wales, 2016). Miller (1983) views entrepreneurial orientation as a compound construct that has three interrelated dimensions: innovativeness, proactiveness and risk-taking. Innovativeness can be defined as ‘the ability of the firm to introduce new products and services or modify existing ones in order to meet the demands of current or future markets’ (Zahra & Covin, 1995). Proactiveness, however, refers to ‘the tendency of the firm to introduce new products and services ahead of the competition and act in anticipation of future demand’ (Wang & Altinay, 2012). As for risk-taking, it refers to ‘the propensity of the firm to commit resources to projects with unknown outcomes’ (Wiklund & Shepherd, 2005). Consequently, in order to draw a comprehensive picture of this phenomenon and obtain more accurate results when investigating entrepreneurial orientation, empirical studies have to take these three components into account (Anderson et al., 2009; Covin & Wales, 2019).

In view of the above, it is expected that companies displaying high levels of entrepreneurial orientation will enhance the effectiveness of strategic decision-making for three reasons. First, companies that adopt innovative ideas have the potential to make more special and competitive strategic decisions, such as creating new product lines or advancing into new markets (Shu et al., 2019; Walter et al., 2006). Second, proactive and risk-taking companies tend to be in an advantageous position due to their pursuit of promising opportunities that allow them to make competitive strategic decisions (Yang & Yu, 2022). Second, proactive and risk-taking companies tend to be in an advantageous position due to their pursuit of promising opportunities that allow them to make competitive strategic decisions. Such decisions could be offering competitive prices, withdrawing present products and offering new products, or starting new investments. This is enhanced by previous studies that suggest that those who enter the marketplace first can gain a competitive advantage (Kimura, 1989). Third, proactive firms are in an advantageous position to react rapidly, since these companies are able to make both prompt and effective strategic decisions because their strategic logic looks for opportunities (Eisenhardt & Martin, 2000; McGee & Peterson, 2019).

Subsequently, these firms are competitively advantageous via the identification of opportunities and their rapid integration into their internal activities. This means that such firms may gain competitive advantages in dynamic business environments when their dynamic abilities match with their strategies (Di Benedetto & Song, 2003; Harreld et al., 2007). In this context, Gumusluoglu and Acur (2016) point out that competitive advantages are often temporary in dynamic business environments. Thus, competition between firms is often nurtured by creating a series of short-term advantages and proactively involving them in their internal activities, rather than establishing a long-term position relating to a given product or technology. Several studies provide empirical evidence to support these arguments (Balabanis & Katsikea, 2003; Covin & Miller, 2014; Knight, 2001; Sundqvist et al., 2012; Wales et al., 2019). In view of the above, this research suggests that:

H₂: The stronger the entrepreneurial orientation is, the more effective the strategic decision-making will be (see Figure 1).

OPEN IN VIEWER Figure 1.

The Conceptual Model with the Research Hypotheses.

Procedural Rationality and the Effectiveness of Strategic Decision-making

Some empirical evidence suggests strong links between procedural rationality in strategic decision-making and strategic decision-making effectiveness, especially when decisions are made to manage crises rather than take advantage of opportunities (Hakeem, 2020; Hall & Van Ryzin, 2019; Pollanen et al., 2017). Dean and Sharfman (1996) highlight a strong relationship between procedural rationality in strategic decision-making and the effectiveness of strategic decisions, especially in unstable environments. Similarly, Elbanna and Child (2007a) reveal stronger links for strategic decisions taken by decision-makers to address crises rather than decisions made to take advantage of opportunities. Previous studies, like Bourgeois and Eisenhardt (1988) have also found a positive relationship between the level of rationality and companies’ performance in environments that are constantly changing; a number of previous studies have displayed similar results, such as (Alarifi et al., 2019; Covin et al., 2020; George & Desmidt, 2018; Glick et al., 1993; Martins & Perez, 2020; Shepherd, 2014). After examining the overall results, it is obvious that procedural rationality in strategic decision-making remains ambiguous, especially in terms of its relationship with strategic decision-making effectiveness. Deeper research is therefore crucial to address the contradictions between prior studies. Taking these points into consideration, this paper suggests that:

H₃: The higher the procedural rationality is, the more effective the strategic decision-making will be (see Figure 1).

Sample and Procedures

The study setting for this paper is a number of companies in Saudi Arabia that belong to Saudi Arabia’s financial sector; namely, banks, finance companies, insurance companies, investments companies, real estate development companies, and Awqaf (endowments) companies. These firms, whose activity is mainly financial, employ more than 100 employees. In terms of collecting the research data, electronic surveys were selected to survey a large and widely-dispersed population. These provide accurate information via the application of one approach on all respondents (Craig & Douglas, 2005; Hair et al., 2008). Surveys have traditionally been used in previous studies (i.e., Elbanna & Child, 2007b; Haniffa & Hudaib, 2007; Jansen et al., 2013; Miller, 1983) as methods of examining procedural rationality, entrepreneurial orientation and strategic decision-making effectiveness. As such, a survey questionnaire was prepared focussing on the aforementioned areas. In total, 1,500 questionnaires were distributed to 375 companies over a period of approximately three months, from mid-December 2019 until mid-March 2020. Out of the distributed questionnaires, 379 were returned from 177 companies. At least two responses from each company were collected so as to guarantee the validity of the research results. However, 15 questionnaires were excluded for various reasons, such as incomplete schedules or irrelevant answers. The remaining questionnaires, 364 from 162 companies, represented a final response rate of 25%, which is quite high and satisfactory.

An initial meeting was held with top management teams, promising to anonymize the research responses and offer a free summary of the results once the survey was completed. Then, the link to the electronic questionnaire was forwarded to the members of the top management team of each firm directly or via the Human Resources Manager. Several measures were taken to avoid participants’ bias. For instance, the research questions were repeated on different pages of the lengthy questionnaire to ensure that the respondents do not link the variables under study (Miller, 2008). In addition, the multiple responses procedures, suggested by Elbanna and Child (2007b), were implemented by asking at least two executives from each company to answer the same questionnaire. Comparisons between the responses of these executives were subsequently drawn. Similarly, this study implemented the Structural Equation Modelling technique to analyse the data.

The study focused on testing causal relationships that were suggested as a research hypothesis to investigate multiple independent and intervening variables. SPSS (Statistical Package for the Social Sciences) (Version 24), and the SmartPLS 3.0 (Partial Least Squares Path Modelling) software packages were used to analyse the collected data.

Even though previous studies have opted for a covariance-based SEM (CBSEM) approach, this study used the PLS path modelling for a number of reasons. First, PLS path modelling is easier in handling formative constructs compared to CBSEM. Although the tools used in CBSEM are capable of handling formative constructs within a study model (i.e., AMOS), only a small number of studies actually adopt such a model, implying there are difficulties in using it (Hair et al., 2016). Second, for research that is exploratory in nature (such as this one), PLS path modelling has been identified as being more appropriate for exploratory research, whereas CBSEM is more applicable and suited to theory testing (Fornell & Bookstein, 1982). Third, there are no restrictive assumptions placed on the data when adopting the PLS path modelling; for instance, PLS path modelling can support reflective and formative constructs, small samples that are less than 100, constructs with single-item, metric and non-metric data types, datasets with multi-collinearity and data sets with missing values (Hair & Alamer, 2022). In light of this, PLS path modelling is recognized as being a ‘soft-modelling’ method.

Research Sample Profile

As previously stated, the sample for this study comprised 177 companies within Saudi Arabia’s financial sector, which includes banks, finance companies, insurance companies, investments companies, real estate development companies and Awqaf (endowments). In relation to the data, the initial data that was returned and collated were from 379 decision-makers as part of the upper-level management of these companies. However, upon further analysis, 15 of the questionnaires returned had to be excluded due to a number of reasons, such as incompletion or non-relevant responses. Thus, the final data sample was 364 participants, representing 164 companies. What follows is a breakdown of the participants’ profile, with regard to their gender, age, level of education, position held and subsequent experience in their position, as well as the type, size, and age of the organization they work for (see Table 1).

OPEN IN VIEWER

Table 1.

Profiles of the Participants and Companies.

Description	Frequency	%
Gender of the Participants
Female	157	43
Male	207	57
Total	364	100
Age of the Participants
25–35 years	51	14
36–45 years	157	43
>46 years	156	43
Total	364	100
Educational Level of the Participants
Doctorate degree	25	7
Master’s degree	180	50
Bachelor’s degree	154	42
Diploma degree	2	1
Other degrees	3	1
Total	364	100
Position of the Participants
Managerial (e.g., strategy, general manager)	188	52
Executive (e.g., CEO, EFO, COO, CSO)	127	35
Board member	19	5
Unknown	30	8
Total	364	100
Sectors of the Companies
Investment companies	38	23
Real-estate development companies	37	23
Finance companies	25	15
Awqaf (endowments) companies	24	15
Insurance companies	23	14
Banks	15	9
Total	162	100
Company Size
<500 employees	77	21
501–1,000 employees	112	31
1,001–5,000 employees	131	36
>5,000 employees	44	12
Total	364	100
Age of the Companies
<10 years	90	25
11–20 years	111	31
21–30 years	33	9
31–40 years	39	11
>40 years	91	25
Total	364	100

The results from Table 1 show that, in terms of gender, male participants made up 57% of the respondents, while 43% were female. This highlights an interesting, yet anticipated, representation of the Kingdom of Saudi Arabia, which, as a result of current reformations, has been identified as one of the five most improved countries in bridging the gender gap within the business sector (Ministry of Labour—Kingdom of Saudi Arabia, 2022). In relation to age, all of the participants fell within 25–63 years old, with 14% in the 25–35 age bracket and 43% in both the 36–45 age group and those above 45. These findings look to be consistent with the country profile, whereby the majority of the population in Saudi Arabia is aged between 20–80 years-old (General Authority for Statistics-Kingdom of Saudi Arabia, 2022). As for education, the results showed that very few were at the diploma or other degrees level (1% for each of these categories), while the majority held a Master’s (50%), followed by an undergraduate degree (42%); only 7% had a Doctoral degree. These findings evidently highlight that those in the upper-level management Saudi financial sector do hold a certain level of academia. In terms of their position/job titles within the top management teams of their respective companies, the findings showed that the majority of the participants were at a managerial position (52%), while 35% identified themselves at the CEO level and only 5% at board level. It is important to note that all the participants were in positions that allowed them to provide comprehensive information on the strategic decision-making actions of the companies they were affiliated with. Furthermore, it should be noted that the approach of sampling different members of upper management, as opposed to only targeting CEOs, is consistent with previous studies on strategic decision-making (Dean & Sharfman, 1993b, 1996; Elbanna & Child, 2007b; Elbanna & Child, 2007a; Goll & Rasheed, 2005), as this improves the generalisability of the research findings.

A further insight and cross-reference of the participants’ education level and position were conducted (Table 2). These findings show that at the executive and board level, the majority of these individuals held a masters’ degree, while for those at the managerial level—a position that requires high-level qualifications—over half held a bachelor’s degree (55%). Table 3 provides a Chi-square test on the above data, indicating a significant different between the education background when compared to the job position within upper management (sig < 0.01, df = 12), which denotes that the executives obtain the highest level of education in comparison to the others.

OPEN IN VIEWER

Table 2.

Educational Level * Position (Cross-tabulation).

	Unknown (%)	Managerial (%)	Executive (%)	Board Member (%)
Diploma	0	1	0	5
Bachelor’s degree	23	55	32	16
Master’s degree	70	40	57	63
Doctorate	7	3	11	16
Other	0	2	0	0
Total	100	100	100	100

OPEN IN VIEWER

Table 3.

Chi-square Tests.

	Value	df	Asymptotic Significance (2-sided)
Pearson Chi-square	44.203a	12	0.000
Likelihood ratio	42.111	12	0.000

The data collection approach adopted in this study was a convenient sampling from a specific sector (i.e., Saudi Arabia’s financial sector), which further enhances the generalisability of the research findings. As shown in Table 1, these sectors consist of banks, insurance companies, Awqaf (endowments) companies, finance companies, real estate development companies and investment companies. The findings were relatively broad across each of these six categories, with the highest respondents coming from both investment companies and real estate development companies (23%), while finance companies and Awqaf (endowments) companies were both at 15%, followed by insurance companies at 14% and lastly, banks at 9%. One of the reasons banks may have had a lower response rate than the other categories is because, as identified by the Saudi Arabia Monetary Authority (2022), there only has 12 main local banks within Saudi Arabia. Company size, as based on the number of employees, was another key component that was collected in the data sample. As highlighted in Table 1, 36% of the companies had between 1,001 and 5,000 employees, whereas 31% had between 501–1000. Hence, the majority of the organizations would constitute medium-sized and above, enhancing the generalisability of the research’s findings. As for those companies that had more than 5000 employees, this was significantly lower in comparison, representing 12% of the data set. Lastly, in relation to the companies’ age, Table 1 illustrates that the majority of the companies in the sample were between 11–20 years-old (31%), with 25% of companies either <10 years-old or over 40 years-old. Only 9% were between 21–30 years-old. This shows that the majority of the organizations in the data sample have had over 10 years of practical experience in their respective area of expertize. This, in turn, enhances the reliability of the responses to the study surveys and thus, enhances the generalizability of these findings.

Independent Variable: Procedural Rationality in Strategic Decision-making

In the present research, procedural rationality in strategic decision-making was examined via a five-item scale taken from Dean and Sharfman (1996) and scaled on a seven-point Likert scale. This measure and its variants have been widely used with similar reliability estimates (Dean & Sharfman, 1996; Elbanna & Child, 2007a; Papadakis, 2006; Thanos et al., 2017). The Cronbach’s alpha value in this study is (0.845), as displayed in Table 4, which is similar to that of previous studies that used a similar scale for measuring procedural rationality (Deligianni et al., 2016; Hakeem, 2020). To sum up, in the current survey, this five-item measure shows acceptable levels of reliability (α = 0.845; CR = 0.890) and validity (AVE = 0.619), as demonstrated in Table 4. Table 5 provides the items, source, measurement and descriptive statistics for the procedural rationality in strategic decision-making construct.

OPEN IN VIEWER

Table 4.

Reliability and Validity Test.

Construct	No. of Questions	Cronbach’s Alpha	Composite Reliability (CR)	Average Variance Extracted (AVE)	DV/IV*
Procedural rationality	5	0.845	0.890	0.619	IV
Entrepreneurial orientation	9	0.923	0.936	0.620	Med
Strategic decision-making effectiveness	4	0.837	0.891	0.671	DV

Note: *DV = Dependent variable, IV = Independent variable, Med = Mediator variable.

OPEN IN VIEWER

Table 5.

Items, Source, Measurement and Descriptive Statistics of Procedural Rationality in Strategic Decision-making.

Construct	Code	Items	Source	Measurement Scale		Mean	SD
Procedural Rationality	PR1	How extensively does the top decision-making group in your firm look for information before making strategic decisions?	Dean and Sharfman (1996)	7-point Scale	(1 = Not at all, 7 = Extensively)	5.62	1.01
	PR2	How extensively does the top decision-making group analyse the relevant information for a decision before making a strategic decision?				5.62	1.041
	PR3	How important are quantitative analytic techniques (such as net present value or discounted cash flow analysis, etc.) in making the strategic decision?				5.51	1.225
	PR4	In general, how effective the top decision-making group in your firm at is focusing its attention on relevant information and ignoring irrelevant information for decisions?			(1 = Not at all effective, 7 = Very effective)	5.52	1.016
	PR5	How would you describe the process that hasthe most influence on the top decision-making group’s decisions?			(1 = Mostly analytical, 7 = Mostly axiomatic)	5.62	1.164

Dependent Variable: Strategic Decision-making Effectiveness

In this research, strategic decision-making effectiveness was calculated using a four-item scale taken from Jansen et al. (2013) and scaled using the same seven-point Likert scale. The Strategic Decision-making Effectiveness variable was measured as the total of these four items. This four-item measure shows acceptable levels of reliability (α = 0.837; CR = 0.891) and validity (AVE = 0.671). Table 6 provides the items, source, measurement and descriptive statistics for the strategic decision-making effectiveness construct.

OPEN IN VIEWER

Table 6.

Items, Source, Measurement and Descriptive Statistics of Strategic Decision-making Effectiveness.

Construct	Code	Items	Source	Measurement Scale	Mean	SD
Strategic Decision-making Effectiveness	SDME1	To what extent have the strategic decisions contributed to the turnover growth of your firm … (1) Not at all effective (7) Very effective	Jansen et al. (2013)	7-point Scale	5.27	1.098
	SDME2	To what extent have the strategic decisions contributed to the profit growth of your firm … (1) Very effective (7) Not at all effective			5.31	1.137
	SDME3	To what extent is the decision-making team satisfied with the decision … (1) Not at all (2) Completely			5.04	1.115
	SDME4	To what extent has the decision led to the expected result … (1) Not at all effective (7) Very effective			5.08	1.011

Mediator Variable: Entrepreneurial Orientation

In this research, entrepreneurial orientation was measured via a nine-item scale taken from Covin and Slevin (1989). Entrepreneurial orientation has three dimensions (innovativeness, risk-taking and proactiveness). This scale has been validated well and is widely used in entrepreneurial orientation research (Edmond & Wiklund, 2010; Rauch et al., 2009; Thanos et al., 2017; Vu, 2017). The Cronbach’s alpha value in this study is (0.837), which is homogenous with previous studies that used a similar scale to measure entrepreneurial orientation (Thanos et al., 2017). In short, the three-item measure in this study showed sufficient levels of reliability (α = 0.837; CR = 0.891) and validity (AVE = 0.671). Table 7 provides the items, source, measurement and descriptive statistics for the entrepreneurial orientation construct.

OPEN IN VIEWER

Table 7.

Items, Source, Measurement and Descriptive Statistics of Entrepreneurial Orientation.

Construct	Code	Items	Source	Measurement Scale	Mean	SD
Entrepreneurial Orientation	EO1	In general, the top managers of my firm do not favour … (1) A strong emphasis on the marketing of tried-and-true products or services. (7) A strong emphasis on technological leadership, and innovations.	Covin and Slevin (1989)	7-point scale	4.79	1.451
	EO2	How many new lines of products/services has your firm marketed in the past 5 years? (1) No new lines of products/services. (7) Many new lines of products/ services.			4.96	1.406
	EO3	Changes in product or service lines … (1) Have been mostly of a minor nature. (7) Have usually been quite dramatic.			4.65	1.483
	EO4	In dealing with its competitors, my firm … (1) Typically responds to actions which competitors initiate. (7) Typically initiates actions which competitors then respond to.			4.81	1.431
	EO5	In dealing with its competitors, my firm … (1) Is very seldom the first business to introduce new products/services, new products/services, administrative technologies, etc. (7) Is very often the first business to introduce administrative techniques, operating techniques, operating technologies, etc.			4.72	1.354
	EO6	In dealing with its competitors, my firm … (1) Typically seeks to avoid competitive preferring a ‘live-and-let-live’ posture. (7) Typically adopts a very competitive, ‘undo-the-competitors’ posture.			5.02	1.354
	EO7	In general, the top managers of my firm have … (1) A strong proclivity for low-risk projects (with normal and certain rates of return). (7) A strong proclivity for high-risk projects (with chances of very high returns).			4.15	1.401
	EO8	In general, the top managers of my firm believe that … (1) Owing to the nature of the environment, it is best to explore it gradually via timid, incremental behaviour. (7) Owing to the nature of the environment, bold, wide-ranging acts are necessary to achieve the firm’s objectives.			4.57	1.38
	EO9	When confronted with decision-making situations involving uncertainty, my firm … (1) Typically adopts a cautious, ‘wait-and-see’ posture in order to minimize the probability of making costly decisions. (7) Typically adopts a bold, aggressive posture in order to maximize the probability of exploiting potential opportunities.			4.38	1.408

Company Size

This research studies the influence of company size based on the number of full-time staff (García-Villaverde et al., 2013; Thanos et al., 2017; Wales et al., 2015). Company size affects its growth and its outcomes (Dimitratos et al., 2004).

Once the Average Variances Extracted (AVE) has been calculated in Table 4, the discriminant validity is to be analysed using the following measures: (a) the Fornell-Larcker criterion, (b) Cross-Loadings and (c) the criterion of Heterotrait–Monotrait Ratio. With regard to the Fornell-Larcker Criterion, Fornell and Larcker (1981) hypothesized that there should be a correlation between each latent variable and its indicator. Thus, to ensure discriminant validity, the AVE of each construct ought to be higher than the latent variable’s highest squared correlation with any other latent variable. For this study, the AVE values of all latent variables were: Entrepreneurial Orientation is 0.788, Procedural Rationality is 0.787, and Strategic Decision-making Effectiveness is 0.819. These values all show that the self AVE for each latent variable is greater than the other variables, as illustrated in Table 8.

OPEN IN VIEWER

Table 8.

Construct Correlation: The Fornell–Larcker Criterion.

Construct	EO	PR	SDME
EO	0.788
PR	0.400	0.787
SDME	0.402	0.487	0.819

With regard to the cross-loading test, the loading of each item was shown to be greater than all of its cross-loadings (see Table 9). In addition, the loading for items over 60% were acceptable results and attained the required cut-off point of factor loading. Thus, the items emerged with their original variables, with no overlapping between items.

OPEN IN VIEWER

Table 9.

Factor Loading for Each Item.

Items	EO	PR	SDME
EO1	0.769
EO2	0.790
EO3	0.858
EO4	0.838
EO5	0.676
EO6	0.798
EO7	0.723
EO8	0.850
EO9	0.768
R1		0.788
R2		0.868
R3		0.791
R4		0.801
R5		0.675
SDME1			0.857
SDME2			0.819
SDME3			0.819
SDME4			0.780

In relation to the criterion of the Heterotrait–Monotrait Ratio (HTMT), Henseler et al. (2016) concluded that, when using this technique to measure discriminant validity, if the value of the HTMT is lower than 0.90, discriminant validity is established between two reflective constructs. For this study, all the HTMT values for the constructs were lower than 0.90, which denotes that the formative constructs were valid (see Table 10).

OPEN IN VIEWER

Table 10.

The Heterotrait–Monotrait Ratio (HTMT).

Construct	EO	PR	SDME
EO
PR	0.437
SDME	0.449	0.566

With regard to the criterion of Inner VIF Values for this study, all the values were higher than 10, which denotes the presence of harmful collinearity (see Table 11). In contrast, any VIF values significantly higher than 1 denote multicollinearity. To surmise, the analysis of the reliability and validity tests indicate that the validity and reliability of the reflective measurement of latent variables were confirmed for all indicators.

OPEN IN VIEWER

Table 11.

The Inner VIF Values.

Construct	EO	PR	SDME
EO			1.191
PR	1.074		1.191
SDME

Mediator Analysis

Figures 2 and 3 shows a model that was provided by Hair et al. (2022) explaining the systematic mediator analysis process found in PLS-SEM. Consequently, this study decided upon the indirect effect of mediation and looks at what type of mediation occurs.

OPEN IN VIEWER Figure 2.

Example of a Simple Mediator Model.

OPEN IN VIEWER Figure 3.

Explanation of the Systematic Mediator Analysis Process in PLS-SEM.

Looking at the results from Table 12 and based on the p values of the effects, the following can be articulated: PR → EO (P1) and EO → SDME (P2) are statistically significant at 0.000% level of significance. Therefore, the requirements of the 1st step of mediator analysis for P1. P2 were met. PR → SDME (P3) was also statistically significant at a 0.000% level of significance. Therefore, the requirements of the 2nd step of mediator analysis for P3 were met. PR → EO → SDME also are statistically significant at the 0.000% level of significance. Therefore, the requirements of the 3rd step of mediator analysis for P1. P2 were met.

OPEN IN VIEWER

Table 12.

Total Effects.

No.	Relationships	Sample	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	P Values	Results
H 1	PR → EO	0.328	0.333	0.045	7.228	.000**	Supported
H 2	EO → SDME	0.246	0.249	0.055	4.441	.000**	Supported
H 3	PR → SDME	0.389	0.385	0.056	6.931	.000**	Supported
H 4	PR → EO → SDME	0.081	0.083	0.023	3.522	.000**	Supported

Notes: **Significant at p < .01.

As highlighted earlier, this paper is based upon previous studies (De Clercq et al., 2015; Dimitratos et al., 2004; Liu et al., 2011; Thanos et al., 2017) by investigating the effects of the aforementioned variables (type of sector, company age and company size) on entrepreneurial orientation, procedural rationality in strategic decision-making and strategic decision-making effectiveness. In terms of company age, the findings in Tables 13 and 14 revealed that the effect of a company’s age on entrepreneurial orientation attained a p value of .000. Since the p values attained were statistically significant at a .1% level of significance, this denotes a positive impact between company age and entrepreneurial orientation. That said, in terms of the effect of the company’s age on procedural rationality in strategic decision-making, the findings revealed the p values for this were higher than .05% (at 0.369), which denotes no statistical significance and therefore, a negative impact of company age on procedural rationality in strategic decision-making. In terms of company size and its effect on entrepreneurial orientation and procedural rationality in strategic decision-making, the findings showed that the p values for both were .000, which is <0.01% level of significance and therefore confirming that company size has a positive effect on entrepreneurial orientation and procedural rationality in strategic decision-making. Lastly, for the type of sectors, the findings showed that, for entrepreneurial orientation, the p values were greater than .05% significance (at 0.132), which means there was no statistical significance and thus, the type of sectors had a negative effect on entrepreneurial orientation. However, for procedural rationality in strategic decision-making, the p value was .001, which denotes a statistical significance and therefore a positive impact of the type of sector on procedural rationality in strategic decision-making.

OPEN IN VIEWER

Table 13.

Confidence Intervals Bias Corrected.

No.	Relationships	Sample	Sample Mean (M)	Bias	2.5%	97.5%
1	Co Age → EO	–0.203	–0.200	0.003	–0.298	–0.098
2	Co Age → PR	–0.040	–0.040	0.000	–0.133	0.041
3	Co Size → EO	0.356	0.353	–0.003	0.252	0.449
4	Co Size → PR	0.226	0.228	0.002	0.122	0.332
5	EO → SDME	0.246	0.249	0.003	0.132	0.348
6	PR → EO	0.328	0.333	0.005	0.238	0.409
7	PR → SDME	0.389	0.385	–0.004	0.283	0.498
8	Sectors → EO	–0.070	–0.070	0.000	–0.156	0.028
9	Sectors → PR	–0.177	–0.177	–0.001	–0.273	–0.073

OPEN IN VIEWER

Table 14.

Total Effects of Control on Procedural Rationality, Entrepreneurial Orientation.

No.	Relationships	Sample	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	P Values
1	Co Age → EO	–0.203	–0.200	0.050	4.039	.000**
2	Co Age → PR	–0.040	–0.040	0.045	0.899	.369
3	Co Size → EO	0.356	0.353	0.052	6.893	.000**
4	Co Size → PR	0.226	0.228	0.053	4.229	.000**
5	Sectors → EO	–0.070	–0.070	0.046	1.509	.132
6	Sectors → PR	–0.177	–0.177	0.054	3.284	.001**

Notes: **Significant at p < .01.

Model Fit

As this study adopted the PLS path modelling, variance explained (R2) is used to assess the study model fit. A number of fairly-recent studies have added supplementary assessment measures of model fit, such as the goodness of fit (GoF) (Henseler & Sarstedt, 2013; Vinzi et al., 2010; Wetzels et al., 2009) and predictive relevance (Stone-Geisser’s Q2) (Hair Jr et al., 2022). Once the validity and reliability of this study’s measurement model were explained, the conceptual model was then assessed and tested via the path coefficients and the explained variance (R2) of the dependent variables. R2 specifies the in-sample explanatory power of the model and the values for R2 range from 0 to 1, in which higher values signify higher levels of explanatory power (Hair et al., 2022; Shmueli & Koppius, 2011; Shmueli, 2010). In terms of the effect size of R2, Cohen (1988) explains that it has a large effect if the value is 0.26, medium if it is 0.13 and small at 0.02. As shown in Table 15, both entrepreneurial orientation and strategic decision-making effectiveness achieved a large effect.

OPEN IN VIEWER

Table 15.

Explained Variance for Dependent Variables.

Construct	R Square	Classified
EO	0.266	Large
SDME	0.288	Large

OPEN IN VIEWER

Table 16.

Stone-Geisser’s Q².

Construct	SSO	SSE	Q² ( = 1 – SSE / SSO)
EO	3276.000	2753.840	0.159
SDME	1456.000	1182.880	0.188

As illustrated in Figure 4, the structural model for this study significantly explains that, for the SDME construct, there is a 28% variation, whereas, for the EO construct, it explains a 26.6% variation. Furthermore, in relation to the predictive relevance (Stone-Geisser’s Q2), Chin (1998) explains that a Q² value that is higher than 0 denotes predictive relevance. The results of these are found in f 16, which shows that entrepreneurial orientation achieved a Q² value of 0.159 while strategic decision-making effectiveness achieved a value of 0.188; this means the study model shows a high predictive relevance and an acceptable fit.

OPEN IN VIEWER Figure 4.

The Structural Model with Estimated Parameters.

In terms of the goodness of fit (GoF), Vinzi et al. (2010), defines this as the global fit measure, which is the geometric means of the R2 of the endogenous (dependent variable) and AVE. The purpose of this measure is to account for the study model at the levels of structural model and measurement, focusing on the overall performance of the model (Henseler & Sarstedt, 2013; Vinzi et al., 2010; Wetzels et al., 2009).

To calculate the GoF, the formula is

G o F = ( R 2 ̄ × A V E ̄ )

The GoF of the measurement model was checked and evaluated accordingly, as shown below:

G o F = ( 0.646 ∗ 0.277 ) = 0.179 = 0.423

The findings showed that the model fit was acceptable. That is, in order to be considered a global valid PLS model, the criteria of the GoF values that are given by Wetzels et al. (2009) are: no fit (<0.1), small fit (0.1–0.25), medium fit (0.25–0.36) or large fit (>0.36). Thus, the value for this model was 0.423, which means it is considered a large fit in the global PLS model validity. Furthermore, to check and evaluate the goodness of fit for the measurement model, the standardized root mean square residual (SRMR) value was used. This measure means that the squared discrepancy between the model-implied indicator and observed correlations should be close to 0 or <0.08 (Henseler et al., 2014). For this study, the findings revealed that the SRMR was 0.065, meaning this study provided support for a model fit.