Factors inFluencing the adoption oF the World Wide Web For job-seeking in south aFrica

In the past decade, the use of the World Wide Web (WWW) as a tool for job-seeking and recruitment has increased globally, changing the dynamics for job-seekers and recruitment organisations. The purpose of this study was to gain greater insight into the factors that influence the adoption of the Internet (WWW) for job-seeking within a South African context. The impact of the Internet (WWW) and newspaper-reading habits on the adoption process was of specific interest. Data was gathered by survey through telephonic interviews with 228 job seekers applying for information technology (IT) work in Cape Town, South Africa. The findings show that the income of a jobseeker influences the favourability of internet facilitating conditions they encounter. Facilitating conditions in turn influence Internet (WWW) usage habits. Such habits influence performance expectancy, effort expectancy and intentions to use the Internet (WWW) for job-seeking. The actual extent of Internet (WWW) usage for job-seeking is positively influenced by these usage intentions and negatively influenced by newspaper-reading habits. These and other findings are discussed and implications drawn.


INTRODUCTION
The Internet has made a wide range of dynamic Web applications available globally, such as electronic mail, online shopping, collaborative forums, as well as online recruitment and job-seeking (Maran 1998).The use of the WWW for job-seeking is growing and has played a major role in introducing new dynamics into the way employers and employees find one another (Kerrin & Kettley 2003).This study aims to develop a better understanding of factors that affect the adoption of the WWW for jobseeking within a South African context.Studying the factors that influence use of the WWW for jobseeking in South Africa and how they shape the adoption process, would add to the body of knowledge on global IT adoption.In South Africa, there are unacceptably high unemployment rates and there is much government focus on employment creation.The outcomes of this study could inform strategies aimed at assisting especially those with no access to mainstream print media to seek employment via public access Internet points.The main question driving this investigation is: 'What are the factors that influence the adoption of the WWW for job-seeking?'The factors investigated are performance expectancy, effort expectancy, trust in the Internet, Internet usage and newspaper-reading habits, facilitating conditions and income.
Also of interest is the potential impact of the Internet's global reach and its potential influence in facilitating migration of skills across national borders.Reasons for studying this phenomenon in South Africa specifically were fourfold.Firstly, there has been a dearth of research on the use of the WWW for job-seeking in South Africa and indeed in Africa as a whole (Mbarika et al. 2005).
Secondly, only 9.6% of the South African population has Internet access (Internet World Stats 2009), hence a small proportion of the employable population.Newspapers are still widely used for job-seeking.It would be of interest to analyse the perceptions and attitudes that would influence the adoption of the WWW for job-seeking.Thirdly, although overall there is a relatively high unemployment rate, there is an acute shortage of skilled professionals in areas such as IT, where the study takes its sample.This could be of interest to recruiting organisations needing to find scarce skills beyond the national borders.The study aims to be useful to both individuals, as well as to organisations as the Internet begins to play a more important role as a job-seeking, as well as a recruitment, tool.In the next section is a general discussion on the WWW and its use for job-seeking.This leads to the establishment of a research framework.The research methodology is then described, followed by a section on the data analysis and results.These results are discussed and implications drawn before the study limitations and ideas for future research are outlined.

INTERNET USE WORLDWIDE
Out of a total global population of 6.7 billion people, there are just over 1.5 billion Internet users (1 596 270 108) representing 23.8% of the global population (Internet World Stats 2009).According to Internet World Stats (2009), Asia had the largest number of Internet users in the world with just over 657 million, representing a penetration of only 17.4%.North America had over 251 million Internet users with a penetration of 74.4%, compared to Europe, which had a penetration of 48.9%.While the number of Internet users in Africa has increased to over 54 million, it had the lowest penetration in the world with just 5.6% of the population accessing the Internet.In South Africa there were about 4.6 million Internet users in 2009, representing just 9.6% of the population (Internet World Stats 2009).While these statistics point to a digital divide in terms of Internet access, there exist other divides in terms of the quality of access, usage patterns and sophistication of use (Dutta et al. 2005).High-speed, low-cost broadband access is prevalent throughout much of the developed world, while in many developing countries reliable broadband access is both rare and costly (Brown et al. 2007).

WWW USE FOR JOB-SEEKING
One of the important uses of the WWW is for job-seeking (Haas et al. 2001;Teo et al. 1999).The number of résumés posted on recruitment websites grew from 100 000 in 1995 to 2.5 million in 1998 and by 2000 the number had grown to 7.65 million (Tong & Sivanand 2005;Simphal 2004).In the United States, about 4 million job-seekers are searching for work on the WWW on a typical day (Barrington & Shelp 2005), searching through approximately 29 million job adverts online, with 2 million new jobs posted monthly (Jansen et al. 2005).According to Simphal (2004), in Germany, half the population uses the WWW and more than 70% of it searches for new jobs by using this medium.In the United Kingdom, 7 million job-seekers used the WWW to look for work in 2003 (Simphal 2004).There are more than 3 million job websites worldwide offering online recruitment services (Simphal 2004).
In South Africa there are several successful online recruitment sites (INet-Bridge 2006).The number of jobs advertised by CareerJunction, one of the leading sites, increased by 60% from 13 060 in January 2005 to 20 982 in January 2006, while over that same period the number of registered job-seekers increased by 44% from 326 641 to 473 313 (Biz-Community 2006).The WWW as a recruiting channel offers several key advantages that encourage adoption by users (Haas et al. 2001).The global reach of the WWW exceeds that of print-based newspapers, making it of benefit to both recruitment companies and job-seekers (Haas et al. 2001).
The ability of the Internet to cross borders has made job-seeking a global activity (Salaff et al. 2004) and has changed the scope for job-seekers, facilitating greater migration (Stevenson 2003).In the United Kingdom, for example, the Internet is actively used to recruit health workers, leading to a significant rise in migration of health professionals from countries such as South Africa, Australia, the Philippines and India to the United Kingdom (Buchan & Dovlo 2004).Chinese professionals and labourers have found jobs in Canada through the WWW (Salaff et al. 2004).Similarly, as Canada benefits from immigration of talent from other nations, it also loses talent to nations such as the United States (Nakamura & Davenport n.d.).
The relatively low cost for posting job positions on company websites, compared to the costs for job adverts in print media, encourages companies to use corporate portals to post job positions (Färber et al. 2003).The Internet offers convenient accessibility for both the organisation and the job-seeker (Liljander et al. 2002).Online communication may also provide for immediate feedback, resulting in a faster hiring cycle and reducing the cost per hire (Galanaki 2002).Some sites also provide additional benefits such as linking job-postings to information about travel, child-care, schools, housing availability and information to help prepare for transition between jobs (Kilder 1999).Some also incorporate video-conferencing to facilitate online interviews (Lievens et al. 2002).In the next section a research framework is presented for this study.

RESEARCH FRAMEWORK
The framework used in this study draws from the Unified Theory of Acceptance and Use of Technology (UTAUT) developed by Venkatesh et al. (2003) and the trust and TAM (technology acceptance model) integrated model of Gefen et al. (2003).The UTAUT framework highlights performance expectancy and effort expectancy as major influences on adoption, as well as facilitating conditions (e.g.infrastructure, support services and training) and social influence.The UTAUT model was derived from a consideration of many other technology adoption models such as the TAM, diffusion of innovations theory and theory of planned behaviour (Venkatesh et al. 2003).It was therefore chosen because of its ability to integrate factors from these other models into one unified theory.Despite its comprehensiveness, UTAUT contains no explicit reference to trust.Trust is an important factor in an online context (Gefen et al. 2003).Mistrust of the Internet could have a negative influence in the adoption of the WWW for job-seeking.Gefen et al. (2003) extended the TAM model by adding the trust concept.In this study, we add trust to the UTAUT model for the same reason.
The adoption phenomenon may be country-or culturespecific due to socio-economic factors (Brown et al. 2006;Srite & Karahanna 2006).Hence in this paper, the effect of income is also investigated.The moderating influence of demographic variables (e.g.age, gender, experience) as suggested by the UTAUT model were not considered in this paper, for the sake of parsimony.Recent research furthermore has shown that as Internet usage gains in global acceptance, the effects of factors such as gender and experience are less pronounced (Zukowski & Brown 2007).

Performance expectancy of the WWW
The UTAUT model recognises performance expectancy as encompassing concepts such as perceived usefulness, extrinsic motivation, relative advantage, outcome expectations and task fit (Venkatesh et al. 2003).In some developing countries, the relationship between performance expectancy and IT usage has not always held true, a finding attributed to differing cultural contexts (Anandarajan et al. 2002).A South African study (Brown & Licker 2003), however, did find a strong relationship between performance expectancy and intentions to use the Internet for learning, providing support for the hypothesis that in this context: H1: Performance expectancy has a positive effect on the intention to use the WWW for job-seeking.

Performance expectancy of the newspaper
Using the WWW as a recruiting tool began in the mid−1980s and it has since grown into a worldwide industry (Galanaki 2002).Contrary to earlier hype concerning the Internet, it has not replaced traditional forms of business, but rather provided an alternative channel by which to reach customers.Job-seekers use numerous avenues to find jobs, including newspapers, employment agencies and the Internet (Haas et al. 2001).The widespread prevalence of newspapers and the historical tradition of finding jobs through newspapers results in heightened perceptions of the usefulness (performance expectancy) of the newspaper for job-seeking.The positive perception of newspapers may in turn negatively influence perceptions of jobseekers around the performance expectancy of the Internet for job-seeking.This may be especially true in developing countries such as South Africa, where the Internet is not widely accessible (World Wide Worx 2005).It is hypothesised therefore that: H2: Performance expectancy of the newspaper has a negative effect on the performance expectancy of the WWW for job-seeking.

Effort expectancy of the WWW
Effort expectancy has been defined as the degree of ease associated with the use of the system (Venkatesh et al. 2003).There have been mixed findings as to the effect of effort expectancy on Internet adoption.These mixed findings have been attributed to the moderating effects of variables such as gender, age, experience, culture and type of task on the strength of relationship (Srite & Karahanna 2006;Venkatesh et al. 2003).In a developing-countries context, due to factors related to cultural values and lack of experience with technology, this relationship has often been found to hold true (Anandarajan et al. 2002).In addition, it has been shown that effort expectancy influences adoption indirectly through performance expectancy (Venkatesh & Davis 2000).This holds true in a South African context too (Brown & Licker 2003).It also has an indirect influence on adoption through trust (Gefen et al. 2003).It is therefore hypothesised that: Vol. 12 No. 1 Page 3 of 9 H3a: Effort expectancy (ease of use) of the WWW has a positive effect on the intention to use the WWW for job-seeking.H3b: Effort expectancy (ease of use) of the WWW has a positive effect on performance expectancy of the WWW for job-seeking.H3c: Effort expectancy (ease of use) of the WWW has a positive effect on trust.

Trust in the WWW
Job-seekers are asked to disclose personal information over the WWW, often to many online job sites.WWW recruitment agencies or hackers could take advantage of online job-seekers through unauthorised use of their personal information.The Internet removes geographic barriers allowing new distant relationships to emerge between recruitment organisations and job-seekers.In these distant relationships, trust in the organisation and the security of the technology are important (Van Slyke et al. 2004).According to McKnight et al. (2002), the degree of trust for a vendor or technology and the uncertainties about security of the medium, such as the perceived risk of hackers stealing personal information, can deter adoption.WWW recruitment sites that rely heavily on personal self-disclosure from the user should take special care of privacy issues.Their ability to build user trust could become a competitive advantage for recruitment sites in the future (Gefen et al. 2003).Users with higher trust levels tend to disclose more sensitive information than users with lower trust levels (Nickel & Schaumburg 2004).The cultural context may also play a part in moderating the strength of relationship between trust and usage intentions (Srite & Karahanna 2006).Trust in Internet commerce has been shown to be of key importance for its adoption (Brown et al. 2006).The hypothesis therefore is: H4: Trust in the WWW has a positive effect on the intention to use the WWW for job-seeking.

WWW usage and newspaper-reading habits
In order to explain the factors influencing usage behaviour, it is important to consider both conscious intentions as well as subconscious habits (Limayem & Hirt 2003).Subconscious habits can be a major factor in influencing technology adoption, since behaviour is often habitual and subconscious rather than intentional (Bagozzi et al. 1992).Behaviour is the recurrence of an action until responses become automatic.The frequency of past behaviour reflects habit and directs future behaviour (Oullette & Wood 1998).
Even though the way jobs are marketed is being transformed by the Internet, recruiters should not ignore traditional media, such as newspapers (Barrington & Shelp 2005).Haas et al. (2001) found that 44% of companies indicated that even though the WWW is a valuable tool for recruitment, the print media is still the primary channel for this purpose.Newspapers are still the preferred job-searching method used by the majority in several countries (Barrington & Shelp 2005;Liang 2005).In developing countries where the WWW is not widely pervasive, the importance of newspapers as a recruitment channel is even more important.
This study explores the general habitual use of the WWW in comparison to newspaper-reading.People who are in the habit of buying the newspaper to read the news or the sport pages are more likely to use the newspaper to look for jobs.They may therefore have more positive perceptions of the performance expectancy of the newspaper.The habitual and frequent reading of newspapers may mean less need to use the WWW for jobseeking, leading to lower intentions to use and actual use of the WWW for this purpose.The hypotheses suggested are: H5a: Newspaper-reading habits have a positive effect on the performance expectancy of the newspaper for job-seeking.H5b: Newspaper-reading habits have a negative effect on the intention to use the WWW for job-seeking.H5c: Newspaper-reading habits have a negative effect on the actual use of the WWW for job-seeking.
Habitual, frequent use of the WWW is associated with intentions to use the WWW for tasks such as banking (Tan & Teo 2000).There is likely to be greater intention to use the WWW for jobseeking too.Actual usage of the WWW for job-seeking is also more likely as a result of this habit (Limayem & Hirt 2003).The hypotheses suggested are: H5d: WWW usage habits have a positive effect on intentions to use the WWW for job-seeking.H5e: WWW usage habits have a positive effect on actual usage of the WWW for job-seeking.
Beyond the influences on intentions and usage, habitual use of a technology is likely to also influence beliefs and perceptions about that technology (Limayem & Hirt 2003).The most common beliefs are those concerning performance expectancy and effort expectancy (Venkatesh et al. 2003).Habitual use of the WWW leads to greater familiarity and self-efficacy.Self-efficacy in turn leads to higher levels of ease of use (effort expectancy) and perceived usefulness (performance expectancy) (Lewis et al. 2003).Trust is just as important a belief in an online context (Gefen et al. 2003).Habitual, frequent use of the WWW should boost familiarity with it.Familiarity in turn increases trust (Gefen et al. 2003).The hypotheses suggested are: H5f: WWW usage habits have a positive effect on performance expectancy of the WWW for job-seeking.H5g: WWW usage habits have a positive effect on effort expectancy (ease of use) of the WWW for job-seeking.H5h: WWW usage habits have a positive effect on trust in the WWW for job-seeking.

Facilitating conditions
Facilitating conditions can be thought of as the organisational and technical infrastructure available to support the users of information technology (Venkatesh et al. 2003).In terms of the WWW, facilitating conditions may refer to the perceived convenience of facilities for accessing the WWW and the sophistication of access method (e.g.some may have WWW access at home or work, using high-speed broadband connections).At the other extreme, some may have access only to public facilities, which may use low-speed, dial-up access.In general, facilitating conditions are most likely to be more favourable in developed countries, due to their sophisticated infrastructure.For example, it was found that in Singapore facilitating conditions were perceived as being more favourable than in South Africa (Brown et al. 2004).These contrasts will have an impact on the ability of users to support their WWW habits.The more favourable the facilitating conditions, the more habitual and frequent the WWW use (Venkatesh et al. 2003).The hypothesis suggested is: H6a: Favourable WWW facilitating conditions have a positive effect on WWW usage habits.
Reading of news and job-searching are common uses of the WWW.The favourability of WWW facilitating conditions may then reduce the dependence on newspapers for such purposes.The hypothesis suggested is: H6b: Favourable WWW facilitating conditions have a negative effect on newspaper-reading habits.Musa et al. (2005) demonstrate that the state of socio-economic development in a region impacts on accessibility and exposure to technologies, which in turn influences their adoption and diffusion.The WWW is a convenient tool for job-seeking but its use may not be spread evenly throughout society (Barrington & Shelp 2005).In South Africa, only about 9.6 % of the population uses the Internet (Internet World Stats 2009).Part of the reason for this low penetration is the high telecommunications costs, which subsequently impact on the cost of Internet access (Gillwald & Esselaar 2004) as well as a communication backbone, which does not extend across all geographical areas.High telecommunications costs in South Africa have kept the Internet out of reach of the majority of the population (World Wide Worx 2005).Sophisticated, high-speed Internet access is more likely among those with higher incomes, suggesting the following hypothesis:

Income and Internet access
H7: Income has a positive effect on favourability of WWW facilitating conditions.
In other words, income has a positive effect on an individual's ability to access the WWW.These hypotheses are illustrated in the research framework in Figure 1 below.

RESEARCH PROCEDURE
A survey strategy was used to collect data via a structured telephonic interview.The telephonic interview approach was chosen because of its natural fit with the interview process of job applicants, given the verbal communication.Telephonic interviews enable a more controlled and consistent way of collecting data by avoiding misunderstandings and spoiled questionnaires in a postal survey.The interviewer is able to ensure that respondents understand the questions and that they respond with due consideration.Telephonic interviews can also produce better quality data and a higher response rate (Venkatesh & Brown 2001).The study used a cross-sectional timeline, with data gathered at one point in time over a limited period in 2006.

Research instrument
The research instrument used for this investigation derived measures for performance expectancy, effort expectancy and facilitating conditions from Venkatesh et al. (2003).The trust measure was based on the work of Gefen et al. (2003).The instrument was tested with a pilot sample over an initial period of two weeks to fine-tune the way the questions were phrased.The number of items per construct was reduced to a minimum through the pilot test in order to keep the length of the telephonic interview as concise as possible.Some items required single quantitative entry (e.g.age, expected monthly income, years of work experience, habits).Other items required a choice from a list (e.g.gender, population group, highest qualification, geographic location, where the job was seen).The main set of measures required respondents to supply a ranked response based on a five-point Likert scale, ranging from strongly disagree (1) to strongly agree (5).The measures for performance expectancy (3 items), effort expectancy (3 items), trust (3 items), facilitating conditions (3 items), intentions to use the Internet for job-seeking (1 item) and actual usage thereof (1 item), employed the Likert scale.

Data collection procedure
The target population was limited to individuals applying for job positions in Cape Town, but regardless of their geographic location, which allowed for a diversity of respondents.It was expected that the majority of respondents for this study would be from Cape Town, since the advertised jobs were in Cape Town.The strategy used to collect the sample data involved the tracking of responses to adverts for real job-positions placed by a recruiting agent.The recruiting agent was aware of the study and agreed to collaborate.The jobs advertised included positions for software developers, system engineers, salesmen and other professions in the IT sector.The adverts were placed with identical wording in popular local Cape Town newspapers as well as on a popular South African Internet job site.The newspaper and the WWW are the channels most commonly used by Cape Town recruiting agents.Applicants Pavon & Brown they may choose to search in newspapers, whereas for jobs in other cities or countries, the WWW may be employed.Some, on the other hand, may have preferences for either the WWW or newspaper channel, regardless of job location.This link between job-seeking behaviour and locality can be investigated in more detail to identify patterns, trends and factors of influence.This research has investigated the phenomenon of WWW adoption from the job-seeker perspective.Future research might investigate what strategies are used by recruitment organisations to maximise the potential of the WWW and newspaper as recruitment channels, taking into account local and international skill availability.Future research might attempt to conduct similar studies in other developing countries to investigate the effect of socio-economic conditions on WWW adoption for job-seeking.Comparative studies can also be conducted, where data can be gathered from several countries and the findings from each compared.
The study was based on a theoretical framework derived from UTAUT (Venkatesh et al. 2003) augmented with a trust variable (Gefen et al. 2003) and measures of newspaper-reading and WWW usage habits.The addition of the trust variable did not add to the explanatory power of the model, as none of the hypotheses associated with trust were supported.Performance expectancy and effort expectancy were confirmed as the major influences on WWW adoption for the purposes of job-seeking.Alternative theoretical models, such as uses and gratification theory, or actor-network theory could also be used to provide deeper insight into the processes of adoption of the WWW for job-seeking.Grounded theory methodology might also be used to uncover aspects of adoption not illuminated through theoretical frameworks such as UTAUT.
Finally, the study has focused on the WWW for job-seeking.The mobile Internet is increasingly pervasive.Future research might investigate the phenomenon of job-seeking via the mobile Internet.

CONCLUSION
The WWW has had a major impact on the way companies recruit.Despite this innovation, traditional ways of recruiting, through newspapers, have not become obsolete.Entrenched newspaperreading habits may be hard to dislodge, making the newspaper a preferred option for job-seeking with some.In developing countries such as South Africa, where WWW penetration is low, use of multiple newspaper and WWW channels may be a necessity.In this study, it was found that use of the WWW for job-seeking is influenced positively by intentions to use the WWW for this purpose and negatively by newspaperreading habits.Intentions to use the WWW for job-seeking are influenced by WWW usage habits and performance expectancy of the WWW.This performance expectancy is influenced positively by WWW usage habits, effort expectancy (ease of use) and negatively by performance expectancy of the newspaper.Facilitating conditions positively influence WWW usage habits and negatively influence newspaper-reading habits.Facilitating conditions were found to be influenced by income.The study has contributed towards an understanding of how habits impact on WWW adoption for job-seeking within a South African context.It has provided a basis for similar studies to be conducted in other countries, using telephonic interviews or alternative data gathering techniques.With increased levels of globalisation and flow of skills across borders, an understanding of job-seeking behaviours and habits is highly relevant.
FIGURE 1Research Framework