Environmental cognitions and scanning behaviour of managers of Chinese small and medium-size enterprise managers: an empirical study of a multidimensional model

Introduction

Environmental scanning research emerged as a new branch of strategic management in 1960s. It refers to

the acquisition and use of information about events, trends, and relationships in an organization's external environment, the knowledge of which would assist management in planning the organization's future course of action (Aguilar, 1967, p. 1).

Although being initiated in management science, until now environmental scanning research has grown up into an interdisciplinary field which attracts attentions of many scholars with different backgrounds, including information science. Information science gradually becomes an important force in the study of environmental scanning, greatly informing that field (Correia and Wilson, 1997; Choo, 2002; Wang, 2004; Alwis, Majid and Chaudhry, 2006; Zhang, Majid and Foo, 2012). The underlying logic of information science participation and contribution is that environmental scanning is actually a kind of executive information seeking and use behaviour.

The target groups of environmental scanning research include not only managers in developed countries, but also those in developing countries. Especially the latter group's scanning behaviour has been of increasing academic interest. So far, however, with few exceptions (Dong, Yan, Liu and Zhang, 2008), there has been little discussion about manager's environmental scanning in China, the biggest developing country in the world. Today small and medium-size enterprises constitute the majority of Chinese economic organizations. 98% of the organizations are of this type and contribute 60% of Chinses Gross Domestic Product. Around half of tax revenue in China is also attributed to these firms. Therefore, the managers of such firms are the critical mass of Chinese managers. Evidence shows that compared with managers of large firms, managers of small companies attach greater importance to environmental scanning and prioritise it (Correia and Wilson, 1997). In order to design and provide better information service to Chinese managers for facilitating company operation and growth, it is important to understand this user group's information seeking behaviour and its influencing factors, The aim of this paper is to critically examine how Chinese managers of small and medium-sized firms scan the external environment and to what kind of degree their contextual perceptions and cognitions affect their environmental scanning behaviour.

Literature review and hypotheses

Managers are ‘sophisticated information seekers' (Boyd and Fulk, 1996, p. 2). They seek information in order to reduce uncertainty (Vandenbosch and Haff, 1997). Indeed, environmental uncertainty is the most studied issue in research on environmental scanning. Perceived environmental uncertainty is defined as the decision-maker's inability to accurately predict environmental changes and confidently assign probability on how environmental factors affect the organization (Duncan, 1972; (Choo, 2002). In information science view, uncertainty is also an essential motivator of information seeking behaviour (Kuhlthau, 2004). A most commonly used indicator of uncertainty in environmental scanning research is perceived strategic uncertainty (Daft, Sormunen and Parks, 1988; Sawyerr, 1993; Elenkov, 1997; Ebrahimi, 2000; Stewart, May and Kalia, 2008). Perceived strategic uncertainty is composed of three dimensions of manager perception: environmental complexity, variability and strategic importance. Environmental complexity refers to the amount and diversity of environmental elements and their interdependent level, whereas environmental variability refers to the rate of change about trends, events and issues taking place in the environment (Duncan, 1972; Boyd and Fulk, 1996; Sawyerr, Ebrahimi and Luk, 2003). Strategic importance describes to what kind of degree the environment is important to attainment of enterprise's goals, implying the extent to which enterprise relies on resources provided by the environment for its performance. Unless the strategic importance of the environment is high, neither complexity nor variability or their combination will give rise to manager's scanning interest and behaviour (Daft et al., 1988).

It is generally believed that the environment cannot be analysed as a whole (see, for example, Bourgeois, 1980; Hambrick, 1979). The appropriate way of dealing with the environment is to classify it into several categories. One classification put forward by Hambrick (1982) is to divide the environment into entrepreneurial, engineering, administrative and regulatory sectors. According to Hambrick, these sectors can be defined as follows:

• the entrepreneurial sector includes trends and events in products and services, demographic trends and shifts of social attitudes affecting demand;
• the engineering sector includes trends and events about optimizing manufacture or logistics of products amd services;
• the administrative sector comprises trends and events of determination of organizational design and maintenance of internal relationships; and
• the regulatory sector refers to government regulations, taxes accreditations, local regulations and litigations involving the industry.

A more popular classification is to decompose the environment into two broad categories: general and task (Fahey and Narayanan, 1986; Daft et al., 1988). The classification criterion is whether the environment influence on business operations and organizational goal attainment is direct. The general environment includes factors having impacts on the organization indirectly. The task environment involves factors interacting with the organization indirectly and influencing daily operations and goal attainment of the organization. The general environment consists of economic, technological, political and legal, and socio-cultural sectors, whereas the task environment consists of customers, competitors and suppliers. We prefer to use the second classification in our research, for two reasons: first, the two-way classification of general and task is the mainstream environmental classification in strategy management research and education (Thompson, Strickland and Gamble, 2010). Since 2000, most environmental scanning research has adopted this classification. It was considered that using the general and task environmental categorization would enable us to compare our results with those of other scholars, leading to a better understanding of Chinese managers' unique characteristics. Secondly, the roles and functions of government and market in Chinese economic development are rather controversial (Zhang, 2010; Lin, 2014)

Managers are assumed to be influenced by these two competing forces in different degrees, generating different-level related uncertainties. By adopting the general and task classification, in which government-related and market-related sectors are explicitly listed and belong to distinct categories respectively, we could separately explore how Chinese managers recognise the two kinds of forces and contrast their relative impacts from a business managers' point of view. It should be noted that, as a matter of fact, in numerous studies technology is categorized as belonging to the task environment (e.g., Daft, el al., 1988; Elenkov, 1997; Stewart et al., 2008). In our study, the technology sector is also included into the task environment on the basis of the fact that 65% of patents and more than 80% of new products in China are produced by small and medium-sized companies (Li and Xue, 2012), indicating technology has become an intrinsic production element of such firms and source of their profits.

Different environmental sectors will cause different perceived strategic uncertainties of managers (Ebrahimi, 2000; Sawyerr et al., 2003). Some studies found that managers in developed countries and areas perceived higher level of perceived strategic uncertainty in the task environment sectors than in general environment sectors (Auster and Choo, 1993; Choo, 2002; Ebrahimi, 2000; Xu, Kaye and Duan, 2003). In contrast, many researches on managers in developing countries like Nigeria, Bulgaria, India and Russia showed that some certain general environment sectors, political and economic sectors in most cases, produced the same or higher level of perceived strategic uncertainty for them as compared with the task environment sectors because of various specificities in those countries such as economic transition , institutional power networks, unstable political climate and so on (Elenkov, 1997; May, Stewart and Sweo, 2000; Sawyerr, et al., 2003; Stewart et al., 2008).

As for the managers of Chinese small and medium-sized companies, the environment they face seems similar to that of developing countries. Although Chinese managers have a great need of information from the task environment (Dong et al., 2008), they may not ignore the complexity and dynamism of the general environment critical to their business. China is undergoing economic reform towards the establishment of a market economy. The macroeconomic environment changes rapidly during the institutional transition. After the new government under chairman Xi Jinping took power in 2013, a new round of deep economic reform was initiated, in which the stock market, monetary policy, international trade, financial and taxation management were involved. These reforms directly influence the business operations of such firms and may generate more strategic uncertainty for their managers than before.

Meanwhile, the economy in China is a Chinese-style market economy because of its peculiar tradition and political system. Despite the fact that private enterprises, mostly consisting of small and medium-sized firms, are acknowledged as the main driving of economy, the entrepreneurs of these organizations still face a lot of bias and threats, the biggest of which is the government monopoly (Coase and Wang, 2012). Chinese governments at different levels exert direct or indirect control of business activities. Child, Tse, and Rodrigues (2013, p. 74) believed that in China ‘although enterprises were given an apparently free hand to manage their affairs, the controlling influence of the bureaucracy was never far away'.

Boisot and Child (1996) pointed out that Chinese governments and government officials at various levels play patronage roles in sanctioning and safeguarding property rights, as well as in acknowledging, supporting and intervening business transactional networks based on Guan Xi, which is the distinctive feature of the Chinese economic order. We have reason to suppose that small and medium-sized firms may keep their close eye on government, and have high uncertainties about them due to their interdependence and government arbitrary decisions. In addition, Chinese central and local governments have intensively enacted laws and regulations promoting the development of such firms in recent years (China. National People's Congress, 2002). Whether managers of these firms can be informed and make best use of these laws and regulations is the key to their business success, implying that it is natural for managers to attach great importance to regulations/legal sectors. Moreover, owing to their volatility and complication, those sectors may create relatively high strategic uncertainty for managers. For the all reasons above, we hypothesise the following:

Hypothesis 1: The rank order of the perceived strategic uncertainties of environmental sectors of managers of small and medium-sized Chinese companies will reflect a mix of task and general sectors.

There has been agreement that perceived strategic uncertainty is a fine predictor of managers' information need (e.g., Daft et al., 1988; Sawyerr, 1993). However, literature has emerged that offers contradictory finds about the relationship between perceived strategic uncertainty and scanning behaviour since the mid-1990s. In their pioneering work, Daft et al. (1988) reported a positive relationship between this uncertainty and scanning frequency for US executives of manufacturing companies. Since then this relationship has been widely examined and has received supported from several studies of managers environmental scanning in different industries and different countries (Sawyerr, 1993; Choo, 1993; Ebrahimi, 2000; Sawyerr et al., 2003). Some of them also found that higher perceived strategic uncertainty increased the scanning interests of managers. Thus, in Daft's opinion, and in the opinion of those influenced by his work, perceived strategic uncertainty will positively affect scanning action and is an adequate predictor of such action.

On the contrary, some in-depth investigations on the scanning behaviour of mangers revealed that perceived strategic uncertainty was not a suitable measurement of manager's environmental cognition and its predictive power was very weak, especially as far as managers in developing countries were concerned (Elenkov, 1997; Wang, 2004). On the theoretical level, the most critical disadvantage of perceived strategic uncertainty is that its components have unique implications and can prompt or prevent scanning in opposite directions (Milliken, 1987). Boyd and Fulk (1996) argued that perceived strategic uncertainty should be decomposed into separate constructs so that the effects of those constructs could be tested individually. They isolated complexity from perceived strategic uncertainty and put forward two variables, perceived complexity and perceived strategic variability (defined as importance × perceived variability). Perceived strategic variability is used to mean the changing rate of trends, issues and events occurring in the strategically important environment. As mentioned before, perceived complexity can be defined as the number and diversity of environmental variables and their interdependence Their empirical study results on the scanning behaviour of seventy-two US executives showed that perceived strategic variability positively influenced environmental scanning, whereas perceived complexity negatively influenced it. Research by May et al. (2000) and Stewart et al. (2008) confirmed the propositions about deconstruction of Boyd and Fulk, and added substantially to our understanding of the why complexity is irrelevant to scanning activity and why importance and variability stimulate scanning frequency. More recently, Dong et al. (2008) also gave Chinese evidence that perceived strategic uncertainty was not an ideal indicator of scanning behaviour. They demonstrated that among Chinese managers, complexity and variability were positively associated with scanning frequency, but at the same time the effect of importance was minimum.

We adopted the cognition measure advocated by Boyd and Fulk (1996), maintaining that to abandon perceived strategic uncertainty and apply perceived complexity and perceived strategic variability may be more appropriate to our current study. The reasons were as follows: first, perceived strategic uncertainty has been approved to be not a useful indicator in predicting environmental scanning for mangers of developing countries such as China, Russia and India; secondly, the variable perceived strategic variability has been seldom mentioned in environmental scanning literature after being put forward, and it is yet quite a good one on the theoretical level. What it needs is validation by a number of empirical studies thirdly, testing the effect of perceived complexity is very common in the research of latter school with Boyd and Fulk as the representatives.

Moreover, managers of small and medium-sized enterprises are ‘gifted readers' of business environment (Smeltzer, Fann and Nikolaisen, 1988 , p. 57). They are sensitive to environmental change. When the external environment changes rapidly, they will intensify their scanning effort because they intend to find opportunities or threats from the change over time. Compared with large firms, this increase in scanning intensity becomes an essential task of the smaller companies by the reason of their vulnerability to fast developing major threats, their lack of resources to absorb the losses related to not recognizing the threats as well as the nature of entrepreneurship, i.e., the ability to discover unexploited opportunities for earning profits (Lang, Calatone and Gudmundson, 1997; Beal, 2000; Stewart et al., 2008). But decision-makers in small firms do not equally allocate their scanning effort to the environmental sectors with the same changing rate because of their limited time and resources. Instead, they are more likely to focus on the important ones of these environmental sectors particularly when such sectors link to organizational performance, strategic advantage, or organizational survival (Boyd and Fulk, 1996). So it may be the case that when managers of Chinese small and medium-sized firms perceive the strategically important environment sectors change rapidly, they will scan those sectors more frequently. Thus we want to test:

Hypothesis 2: The higher the degree of perceived strategic variability, the higher the level of scanning frequency of managers of Chinese small and medium-sized firms.

Perceived complexity has also proved to be an important topic for the environmental scanning community. The trend in the business environment in the 21th century is to become more complex. Just as Boisot (2000) stated, executives today have to address ever more environmental issues and more elusive, nonlinear relationships between them. How do managers response to increasing environmental complexity? When environmental complexity is very high, the manager's ability to analyse and predict the effect of environmental issues on the organization will become low; and the value of information emanating from the environment for decision-making will also be low (Boyd and Fulk, 1996). In this situation where individual manager only has limited abilities to search, process, interpret and use environmental information (Ebrahimi, 2000), s/he may reduce or stagnate scanning effort and turn to rely on his or her own introspective knowledge structure, including intuition, experience and judgment to make decisions (Stewart et al., 2008). In their empirical study, Boyd and Fulk found there existed a negative relationship between perceived complexity and scanning frequency. Taking account of these arguments, we suggest the following:

Hypothesis 3: The higher the degree of perceived environmental complexity, the lower the level of scanning frequency of managers of Chinese small and medium-sized enterprises.

Managers' environmental scanning behaviour is affected by information attributes, the most prominent of which is information source quality. Along with Culnan (1983), in their study of 115 Canadian CEOs' information seeking behaviour, Auster and Choo (1994) found that perceived quality of information sources was a much better variable in explaining use frequency than perceived source accessibility and environmental uncertainty. After eight years, Choo collaborated with Marton to explore the source using behaviour of women information technology professionals in the networked environment. They discovered a significant positive relationship between source quality and source use (Marton and Choo, 2002). Choo (2002) also generalized the findings of information source use patterns in scanning, concluding that perceived source quality certainly influences source preference of managers. From the theoretical perspective of information science, Lancaster (1995) pointed out that, under the circumstances where people have strong motivations to seek information, the quality of an information source will become a determinant factor of source selection. Based on the foregoing, it is possible that a strong link may exist between perceived quality of information source and scanning frequency of managers of Chinese small and medium-sized firms. Furthermore, we expect that these managers will scan more frequently when perceiving an information source to have high quality and sensing a high degree of environmental complexity or strategic variability than they will when only perceiving the high degree of complexity or strategic variability. Hence, it may be hypothesized that:

Hypothesis 4a: The greater perceived quality of information source, the higher the level of scanning frequency of managers.

Hypothesis 4b: As moderator variable, perceived quality of information source will reinforce the relationship between the perceived complexity and scanning frequency of managers.

Hypothesis 4c: As moderator variable, perceived quality of information source will reinforce the relationship between perceived strategic variability and scanning frequency of managers.

Figure 1 illustrates our general research model about the hypothesized relationship between environmental cognition and the scanning behaviour of managers of small and medium-sized Chinese companies.

Method

Sample and data collection

Our sample consisted of managers of small and medium-sized firms (hereafter, simply referred to as 'managers') located in Tianjin, which is one of the four metropolitan areas directly under the Chinese Central Government. We agree with Case's view that managers can be defined as 'individuals who have at least some university education and who work in a sizable organization' (Case, 2002, p. 248). We further added that the lowest requirement of being a manager is that the individual must be a senior official in a firm. The study population came from two groups:the first group was Tianjin managers attending an MBA programme and the post-graduate programme of management science in Tianjin Normal University, which is the largest university in Tianjin. The second group was Tianjin managers who were introduced to us by Tianjin Science and Technology Committee, a department of Tianjin local government. After face-to-face contact or phone calls, a sample of 169 managers agreed to participate in our study.

We used the questionnaire as our data collection tool since it is a common tool and mainstream method of data collection in information science and environmental scanning, and statistically significant relationships can be revealed by it (Alwis et al., 2006). The questionnaires were distributed to the 169 managers. A letter setting out the survey purpose and promising confidentiality of respondents was also enclosed in each mail. Of the 169 managers surveyed, 120 completed and returned usable responses. The overall response rate was 71%, which was relatively high in surveys of senior managers. Familiarity with many of the first group and active personal contacts with second group may contribute to highly effective response rate. More importantly, the dominance of Chinese collectivism culture and traditional Confucian culture with benevolence and propriety as its core in Chinese capitalism (Redding, 1990) is possibly the major factor in such high response rate of Chinese executives. The survey work was carried out over a four month period.

Before data collection, we made a pilot study to test the scale reliability of questionnaire items. The subjects of that pilot study were 23twenty-three Tianjin managers who were not included in the formal survey. The Cronbach alphas (a coefficient of reliability) were acceptable for all items except the technology sector. We revised the wording of the technology item in the questionnaire to make it more precise. Then we asked two Chinese managers, one management professor, two information science professors and three information science post-graduate students to examine the questionnaire. Depending on their suggestions, we modified the questionnaire slightly. In this manner, the content validity of the questionnaire was warranted. Cronbach alpha estimates of items related to the study in the formal questionnaire ranged from 0.670 to 0.856, and the Cronbach alphas of overall scales is 0.943, meaning that the reliability coefficients for all scales in this study were very satisfactory.

The demographic characteristics of the sample are shown in Table 1, below. These managers worked in a variety of industries, which is consistent with the recommendation that industry sample should be multiple (Hambrick, 1981).

Measures

In line with the work of Daft et al. (1988) and later researchers, we measured the perceived strategic uncertainty being composed of three variables: environmental variability, environmental complexity and the strategic importance of the environment. By applying the five-point Likert scale (5=very high, 1=very low), we measured the perceived variability, complexity and strategic importance of six environmental sectors, the item questions of which were designed according to the definitions mentioned earlier. To evaluate the perceived variability, the respondents were asked to answer the question. 'How would you rate the changing speed in the trends, issues and events in each environmental sector?' The respondents were also asked to provide estimates of perceived complexity and strategic importance by answering, 'How would you rate the number and diversity of events taking place in each environmental sector?' and 'How important do you think each environmental sector to be in attainment of your company's goals', respectively. The formula of perceived strategic uncertainty is:

perceived strategic uncertainty=strategic importance × (environmental complexity + environmental variability)

We adopted Boyd and Fulk's (1996) measure to view perceived strategic variability as the product of environmental variability and strategic importance of the environment. We used these formulae to calculate the perceived strategic uncertainty and perceived strategic variability scores for each environmental sector.

To assess perceived source quality variable, we employed two sub-dimensions, i.e., relevance and reliability of information to represent it with reference to Auster and Choo (1994) and May et al. (2000). We also applied the same Likert scale (5=very high, 1=very low) to the measurement of the two items, relevance and reliability of information. The respondents were asked to rate the relevance ('To what kind of degree is the acquired information useful to the corporate operation and goal attainment.') and reliability ('The extent to which the acquired information is authoritative, accurate and truthful.') for each of four information source. The four information sources were impersonal external, personal external, impersonal internal and personal internal respectively. The index of perceived quality of each information source was obtained as a result of summing the value of relevance and reliability. The total perceived quality of information source was the mean value of four kinds of perceived source quality.

As for measurement of scanning frequency, we requested the respondents to evaluate how frequently they access information about six environmental sectors on a five-point scale. The scale ranges from daily (365), weekly (52), monthly (12), quarterly (4) to yearly (1), which is a slight revision of Sawyerr (1993) and Sawyerr, Ebrahimi and Thibodeaux (2000). To eliminate the skewness of the distribution of the dependent variable, we took the natural logarithm of value of scanning frequency (Xie, 2013), resulting in an almost normal distribution of scanning frequency.

Several demographic factors were included in the survey. Similar to Hambrick (1981) and Hambrick and Mason (1984), we gathered data about sex, age, education degree and management position from each manager. Through analysis of the data set, we can find whether these factors impact upon scanning behaviour.

Results

The means, standard deviations of each variable for six environmental sectors and their comparisons with results of other studies are presented in Table 2. As can be seen, the mean values of perceived variability, perceived complexity and environmental importance for each six sectors in our study are generally higher than those for US executives, indicating the both task and general environments for corporate managers were more turbulent and complex in China than in US. And Chinese managers tend to attach greater importance to the environment because of their heavy dependence on external resources and knowledge to compensate for the shortage of internal resources and limited capabilities in pursuing the organizational goals. Moreover, compared with Chinese managers in general, the study managers usually perceive higher levels of environmental complexity and variability and regard the environment as more strategically important. This may be related to the observation by Smeltzer, et al. (1988) of small businesses being gifted readers of the environment: their managers are more sensitive to the subtleties of business ecology.

Hypothesis 1 assumes that the order of environmental sectors according to the perceived strategic uncertainties of Chinese SMEs managers will show a mixed pattern of task and general sectors. We used one-way analysis of variance (ANOVA) to test hypothesis 1. Table 3 showed results (F=24.878, p=0.000 <0.001). From this data, we can note that there were significant differences in perceived strategic uncertainties across the six environmental sectors. We further used Scheffe's multiple range test at a 0.05 significance, which was a well-proven method in environmental scanning research, to examine the statistical differences in perceived strategic uncertainty for each environmental sector. As displayed in Table 3, the sectors ranked in descending order of perceived strategic uncertainty were as follows: customer, technology, legal/regulations, macroeconomic, competitor and sociocultural. The environmental sector with the highest score was customer sector and the second was technology. Being parts of the general environment, regulations and macro economic sectors were in the middle positions of sequence with relatively high scores, followed by competitor sector in the task environment. The top three sectors (customer, technology and legal and regulations) had significantly higher scores than macroeconomic, competitor and sociocultural sectors. The score for the sociocultural sector was significantly lower than all other sectors. Therefore Hypothesis 1 was supported. We also listed the sector rankings of some classical studies such as Daft et al. (1988), May et al. (2000) and Stewart et al. (2008) so that we can compare our results with them.

It should be mentioned that the study by Daft et al. (1988) dates from twenty-seven years ago, which, of course, pre-dates the World Wide Web. We believe, however, that their work is still pertinent today. In fact, it is not uncommon for their work to be cited for comparison in environmental scanning research since Internet became widely used (Sawyerr et al., 2000; Stewart et al. 2008; Dong et al. 2008). The reason we use Daft et al. for the purposes of comparison is that, because although information technology has changed greatly, managers still face and perceive the same features of the environment. Environmental uncertainty, importance, complexity and viability exist objectively whether in the print age or in the digital age. Managers' mindsets, mind models and cognition abilities will be relatively stable whereas the means they access to information become very different as the technology has evolved. They will always perceive these characteristics of the business environment to make decisions about the allocation of scanning efforts. What technology contributes mostly is that it changes the pattern of scanning behaviour, for example, more managers rely on the Web rather than print documents to obtain environmental information.

We employed hierarchical regression analysis to examine Hypotheses 2, 3 and 4. The strength of method lies in its ability to estimate the unique effects associated with each independent variable. This method enables unique partitioning of the total variance of the dependent variable accounted for by correlated independent variables. Each regression in the hierarchical procedure gives a variance increment of the dependent variable, explained by the independent variable entering at that stage (Cohen, Cohen, West and Aiken, 2003). May et al. (2000) and Stewart et al. (2008) have used this method to measure the impact of environmental perceptions of managers on their scanning frequency and scanning modes, gaining satisfactory findings. When applying hierarchical regression analysis, we firstly conducted mean centering of the variables to eliminate the possible effect of multicollinearity between them (Jaccard and Turrisi, 2003). Diagnostics of the regression models in our study showed that tolerance levels and variance inflation factors were satisfactory: tolerance levels ranged from 0.473 to 0.968 and inflation factors from 1.004 to 2.115, indicating the multi-collinearity did not exist among independent variables in our research.

In the next step for hierarchy analysis, we specified the entering sequence of variables organized in terms of our research intention and causal priority. We entered such demographic variables as sex, education degree, age and management position into the analysis at first since they are likely upstream variables of causal flow in 'behaviour research (Cohen et al., 2003). Then we entered perceived strategic variability, perceived environmental complexity and perceived information quality into the equation in turn. According to the advice of Jaccard and Turrisi (2003), we introduced the multiplicative interactions of perceived strategic variability and perceived information quality, perceived complexity and information quality into the equation at last after the main effects. Residuals examination showed there were no violations of regression assumptions. Table 5 presents the regression equations. We found no significant relationship between each of demographic variables and scanning frequency, which was consistent with May et al. (2000) and with research by Stewart et al. (2008), which showed age, position and years of education did not significantly influence scanning frequency, and Hambrick's (1979) observation that there was no consistent relationship between hierarchical level and scanning activities of executives.

Hypothesis 2 predicts that when managers perceive a higher degree of perceived strategic variability, they will scan the environment more frequently. The results demonstrated a statistically significant relationship between that variability and scanning frequency. Being key variable, variability had also the greatest contribution in explaining scanning frequency. Hence Hypothesis 2 was supported. Hypothesis 3 predicts that, when managers perceive a higher degree of environmental complexity, they will decrease the frequency of environmental scanning. The figure in Table 5 told us that this hypothesis was not supported. On the contrary, the relationship between perceived environmental complexity and scanning frequency was significantly positive for our sample.

Hypothesis 4a states that managers will scan the environment more often when they perceive the quality of information source is high. The results provided support for Hypothesis 4a. In Hypothesis 4b and 4c, we expect that perceived quality of information source will positively condition managers' scanning response to both and environmental complexity. Judging from the statistics, we revealed when these managers perceive higher degrees of perceived strategic variability and higher source quality, they will scan the environment more frequently. What is some surprising is that entering the interaction between perceived environmental complexity and perceived quality of information source in step 5 resulted in a significantly negative relationship between the interaction terms and scanning frequency. This indicates as a moderator, perceived source quality weakens rather than strengthens the positive effect of environmental complexity on scanning frequency. So we can draw the conclusion that Hypothesis 4b was supported, and that Hypothesis 4c is not.

Discussion

Our study set out with the aim of examining environmental scanning behaviour of managers of small and medium-sized Chinese companies and the factors that impact upon that behaviour. These managers are a neglected user group in both information science and strategic management research. The results indicate that the studied managers in their unique institutional and economic context exhibit both similar and different features in environmental cognition and scanning as compared with managers in developed countries or other emerging markets. The current study also adds knowledge to the research field of factors influencing environmental scanning, where theoretical debates have arisen in recent years (Sawyerr, et al., 2003; Stewart, et al., 2008) .

As anticipated, environmental sector rankings for perceived strategic uncertainties showed a mixed pattern of task and general sectors in our sample. In this aspect, our managers differ from those in Western, developed countries and are similar to Russian managers (May et al. , 2000). They attribute high uncertainties to both task and general environment. As far as the relative magnitude of perceived strategic uncertainy for each environmental sector is concerned, the customer sector generated the highest score. China has struggled to establish a market economy system. In a mature market economy, the most important decisions entrepreneurs have to make in uncertain situations is to find the relevant price, i.e., predict the price of a product or service that customers are willing to pay. Only by doing that, can they survive and earn profits (Coase, 1937). To conduct such predictions accurately and in a timely manner, managers must attach great importance to customers, study the detailed typology of customers and be aware of any small changes in their needs, preferences and behaviour. Therefore, customers become the main source of environmental uncertainty for managers. This finding is agreement with the findings of Daft et al. (1988), Sawyerr (1993), Auster and Choo (1993), May et al. (2000), Wilson (1994) and Dong et al. (2008), all of which corroborate Aguilar's original idea that 'market tidings predominated as a shared concern of all managers' (Kotler, 1967, p. 538). This also indicates that the Chinese economy has the basic features of a market economy.

The managers in this study recorded the second highest measure of perceived strategic uncertainty in the technology sector. Many small firms in China pursue technology-driven growth. Technological innovation has caused great concern for managers of these firms in recent years. Chinese central and local governments also established a number of policies to encourage research and development activities in such firms. For example, Tianjin municipal government issued the regulation aimed at fostering '"small tech-giants' and set up a special fund to support the development of Tianjin high-tech small firms (Tianjin municipal government, 2010). Under such circumstance, it is natural for managers of this kind of company to pay attention to the technology sector and be alert to the ecological structure and changes in the technology domain. This also accords with the observation byStewart et al. (2008) of U.S. and Indian entrepreneurs.

The political and legal sector generated relatively high strategic uncertainty, showing no significant difference from customer and technology sector on the score for perceived uncertainty. This is not uncommon among investigations into environmental scanning of managers in developing countries, such as those by Sawyerr (1993) and Elenkov (1997). In China, governments are in control of daily operations and development of all business companies (Child, Tse and Rodrigues, 2013). The arbitrariness and opaqueness of government behaviour, together with inefficiency of the legal system, create a high degree of uncertainty, the kind of which maybe the most difficult to address by Chinese entrepreneurs (Zhang, 2010).These things result in a low evaluation of Chinese macro-business environment. The World Bank Doing business report (2013) demonstrates the macro-business environment in China is not so optimistic as some foreigners conceive: China is ranked 96th in the country list of the ease of doing business owing to the problems mentioned above. However, a promising trend is that China is rapidly catching up with the more developed countries through political reform and legalization, as evidenced by the World Bank report.

The economic sector also has high strategic uncertainty, greater even than competitor sector. A possible explanation for this is that the Chinese government is undergoing a new round of economic reform at the current critical juncture of ‘feel the stones when crossing the river', a popular saying by Chinese former leader Deng Xiaopeng to describe the Chinese reform process. The general economic uncertainty and the reform programmes result in the high perceived uncertainty of Chinese managers, especially those of small and medium-sized firms. The social norms and culture change slowly in China largely due to this nation's long-term accumulation of historical customs and the solid roots of Confucian tradition. This is the reason why Chinese managers gave the lowest strategic uncertainy score to the socio-cultural sector.

Managers of the small firms studied in this research responded to increasing perceived strategic variability with higher environmental scanning frequency. This finding confirms the argument of Boyd and Fulk (1996) that perceived strategic variability is an important predictor of managers' scanning frequency. In the Chinese situation, with a deep–level reforming programme and a large domestic market, any change in important environmental sectors often implies a considerable potential of business profit for managers. It is these manager who may be better able to take advantage of such business opportunities since their firms are more flexible than large companies, and s/he is able to formulate or renew the strategy and make the organization adaptive to the environment in time. So s/he has the stronger motivation to seek a greater amount of relevant information about the changing sectors to capture business opportunities, conduct and execute appropriate strategic planning and consequently earn profit.

Contrary to our expectations, the study result statistically supported a positive rather than a negative relationship between perceived environmental complexity and the frequency of environmental scanning activity. There are two possible explanations for this significantly positive relationship.

First, facing the environmental complexity might stimulate highly creative entrepreneurship of Chinese corporate decision-makers, and, as entrepreneurs, tend to devote themselves to creating value in such an extremely challenging situation (Puffer, McCarthy and Boisot, 2010). The entrepreneurial spirit will drive Chinese managers to intensify their environmental scanning frequency and access to more information when the environment becomes more complex. Second, our finding also echoed Ebrahimi's (2000) research into Hong Kong executives' scanning behaviour. We infer that it is the unique mindset of Chinese executives that matters. In fact, strategies of handling environmental complexity are different in China and in western countries. In western countries, managers and their organizations generally reduce complexity through codification and abstraction, for example, by designing routines and standards. In China, however, managers and their organizations prefer to absorb complexity by establishing close-knit relational networks for sharing the risks associated with complexity and ambiguity (Boisot and Child, 1999). By comparison with the complexity reduction strategy popular in the Western, the complexity absorption strategy that is more suitable to cope with uncertainty and complexity offers Chinese companies a higher possibility to renew them and adapt to the environment (Boisot and Child, 1996, 1999). This means that, supported by relational networks like Guan Xi, Chinese managers deploying a complexity absorption strategy may have greater capacity to address the challenge of environmental complexity. Therefore, they appear to increase scanning frequency to explore the higher complexity rather than avoid it.

We also revealed that the higher a manager perceived the quality of information sources tobe, the greater the scanning frequency. This finding added more evidence to the opinion that perceived information quality plays a more important role than other information factors in determining manager scanning behaviour, which was previously put forward and empirically tested by Auster and Choo (1993), Choo (1993) and Marton and Choo (2002). Our finding can be explained from the perspective of domain analytic paradigm (Hjørland, 1997). In the domain analysis theorist's view, the basic research unit in information science is a discourse community. Different discourse communities may have different knowledge structures, communication patterns and relevance judgment criteria. According to HjørlandThe position of discourse community in the structure of social division of labour shapes the information behaviour pattern of that community. Are managers a distinctive discourse community? The answer is 'Yes'. Just as Correia and Wilson (2001) pointed out, managers are special information users behaving in a special information–use context. So managers, as a discourse community distinctive from other communities, may exhibit unique preferences for informational characteristics when seeking information. Because managers are situated in a turbulent and complex business environment where strategic and tactical information is critical to the organizational survival and success, and their social roles formed by intellectual labour division requires them to make unstructured decisions about unpredictable and ambiguous issues (Taylor, 1986; Choo, 2002; Wang, 2004), they will place a premium on the quality of information such as reliability and relevance during environmental scanning process.

We found that the perceived quality of information source significantly moderated the linkage between perceived strategic variability or complexity and scanning frequency. But the directions of the two moderating effects were opposite.

First, compared with managers who only perceive a high degree of strategic variability, those who perceive both high degrees of strategic variability and high quality of information source have a propensity to scan the environment more frequently. This is because, when information sources are of high quality, when managers sense the increasing change rate in important environmental sectors, they will put more efforts to search and use information while, being free of extra cognitive burden of filtering and justifying the scanned information. What more information brings is only better business insights: more relevant and reliable information can provide more possibilities for the manager entrepreneur to creatively discover economic knowledge from information (Boettke, 2002). ‘More is better' works. So we confirm that Chinese managers respond more positively to their perceptions of changes occurring in strategic environmental sectors by scanning when they believe information source is useful, authoritative, accurate and truthful, i.e., of high quality.

Second, contradictory to our expectation, perceived quality of information source negatively conditioned the relationship between perceived environmental complexity and scanning frequency. Our findings meant that on condition that the information source has low-quality, higher level of perceived complexity led to an increase in executives scanning amount. In contrast, when the information source was perceived to have high quality, the effect of perceived complexity on scanning frequency was weakened. We discuss these results by employing analysability and informativeness to feature complexity and quality respectively. A key measure of perceived environmental complexity is manager's perception of analysability for events or trends in environment, which means the extent to which manager can apprehend or describe cause and effect relationships for environmental issues (Boyd and Fulk, 1996).The low level of perceived analysability implies a high level of complexity. And the informativeness that managers need to identify the cause-effect relationship is certain. The valuable information is concentrated in high-quality source, but dispersed in low-quality source. The results can be explained in a twofold aspects. On the one hand, when managers think the analysability is low and perceive the quality of an information source not so good, they will more actively seek a great amount of information because they need to accumulate enough information to extract, synthesize and analyse in order to clarify the cause-effect relationship. On the other hand, a manager seeking to understand causal relationships in the environment, may achieve this with a small amount of information from a high quality source, rather than with much more information from lower quality sources. That is to say, a little high-quality information may function as well as a great amount of low-quality information, so that managers will spend less time and resources on scanning activity than they will in the former situation.

Conclusion

Environmental scanning is an important kind of executive information seeking 'behaviour. It is critical to organizational success and even survival, which is especially true for small and medium-size enterprises. In this paper we examined the relationship between several dimensions of managers' cognition about the external environment and their environmental scanning behaviour in China. The results were that managers ranked the strategic uncertainties of environmental sectors in a mixed pattern. Both perceived strategic variability and environmental complexity positively affected their scanning behaviour as measured by frequency. Our findings also showed that there was a significantly positive relationship between perceived quality of information source and scanning frequency. The Chinese managers intensified scanning frequency when they believe the strategic variability and quality of information source were both high, but decreased that frequency when both environmental complexity and the quality of information source were high.

We have several suggestions for future studies in this field. First of all, environmental scanning research field is like a delightful tourism landscape that many tourists are attracted to visit but nobody resides in. Not few scholars from various disciplines have interests in environmental scanning and contribute their insights to the field. But unfortunately their interests do not last for a long time. Perhaps scholars' enduring concentrations are most needed now for the prosperity of this research field.

Secondly, although various measures of managers' cognitions were used as independent various, only one dimension of scanning behaviour, that is, scanning frequency was used in our research. Scanning frequency describes the volume or quantitative aspect of scanning behaviour, whereas the qualitative aspects of scanning are also very important and worth investigating. The qualitative aspect of scanning is mainly concerned with how satisfied or sufficiently informed managers perceive when they make decisions with the support of scanning activities. It is more closely related to scanning effect and corporate performance and may be a better indicator of scanning behaviour. The real question for the future is to examine the significant relationship of executives' environmental cognitions and the quality aspects of scanning behaviour.

Thirdly, some information science theories and models have been used in environmental scanning research. Nevertheless it is not enough. There is a large potential space for application. Wilson (2008) noted that from a publishing perspective, information science played a dominant role in information behaviour research. It is somewhat regrettable that, until now, in the subfield environmental scanning, information science has not had so great disciplinary influence as management science. What information science scholars should do is to put more effort into applying theories and models of information behaviour to environmental scanning research. In our further research, it is possible investigate how managers' environmental cognitions influence their scanning behaviour process. Ellis's model of information-seeking behaviour, Kuhlthau's information search process and other well-known models in information science need to be introduced into environmental scanning research. As being the key components of the behavioural dimension, these models can be incorporated with environmental scanning theories in management science to empirically develop an integral executive cognitive-behavioural model of information need, seeking and use, which is basically interdisciplinary. The time for close collaboration between information science and management science in relation to environmental scanning is clearly at hand.

Acknowledgements

We are grateful for extremely useful discussions and comments by Professor Liu Chunmao. This paper is an achievement of youth project of Chinese National Social Science Fund "Research on influencing factors and search patterns of SME managers' information 'behaviour" (Number 13CTQ040).

About the authors

Lin Wang is an associate professor in the School of Management at Tianjin Normal University. He received his PhD degree from Department of Information Management Peking University, China. His current research interests include environmental scanning, information 'behaviour and foundation of library and information science. He is the corresponding author for the article and can be contact at wanglinpku@163.com.
Fei Guo is a graduate student in library and information science in Chengdu Library, Chinese Academy of Sciences. She received her Bachelor's degree from the School of Management Tianjin Normal University, China. She can be contacted at 45537298@qq.com.