The roles of LinkedIn-based skill endorsements and LinkedIn-based hiring recommendations on hiring preferences: evidence from Bangladeshi employers

Abstract

Purpose

LinkedIn is a popular and maybe the best job-oriented social networking site (SNS) around the world. Numerous individuals (mostly, jobseekers), as well as the companies (mostly, hiring organizations) have LinkedIn accounts. This study intends to reveal the roles of LinkedIn-based skill endorsements (LSEs) as well as LinkedIn-based hiring recommendations (LHRs) on the hiring preferences (HPs) of Bangladeshi employers.

Design/methodology/approach

The authors purposefully selected 388 valid respondents (employers) from different Bangladeshi job sectors. Based on their responses, the authors applied SPSS 25 for explanatory statistics and a “Covariance-based Structural Equation Modeling (CB-SEM)” (with AMOS 25) to measure the hypothesized associations.

Findings

After appropriate analysis, the authors revealed a momentous positive role of LHR on HP, whereas LSE was identified to have an inconsiderable role on HP.

Originality/value

In this empirical study, the authors tried to highlight the relatively unattended area of Bangladeshi employers' HPs that can be affected by LSEs and LHRs. The authors believe that this study will encourage the researchers to unveil the impacts of LinkedIn-based profile characteristics on job candidates' employability. This empirical study can also aid the employers and policymakers to look for a new avenue for hiring competent talents.

Citation

Hosain, M.S. and Mamun, A.M.A. (2023), "The roles of LinkedIn-based skill endorsements and LinkedIn-based hiring recommendations on hiring preferences: evidence from Bangladeshi employers", Management Matters, Vol. 20 No. 2, pp. 169-184. https://doi.org/10.1108/MANM-05-2023-0021

1. Introduction

After its inauguration in 2003, LinkedIn has now become the most popular job-oriented social networking site (SNS) among the job seekers for searching their preferred jobs, among the employers for searching their expected job candidates and among the individuals (who are already serving) for networking with fellow users as well as for getting their expected career guidance tips and suggestions (Roulin and Stronach, 2022; Khan, 2021; Hosain and Liu, 2020a, b; Roulin and Levashina, 2019). A report from Linkedin.com indicates that by the end of 2022, the site had more than 850m users worldwide and more than 58m registered institutions in more than 200 countries and areas all over the world. Despite the fact that each user's goal may vary, this site is exceptionally popular and appreciated by fresh graduates, newly hired individuals and recruiting firms for its exclusivity. LinkedIn was specifically created and developed with a single purpose to build a professional network combining job searchers, hiring firms and career counselors; unlike other social media (SM) platforms like Facebook, Twitter or YouTube. Therefore, the vision of LinkedIn as an SNS is not to share personal opinions, views or feelings that the users do on other SNSs; rather the users connect through this SNS for the solo purpose of being hired, searching the right candidates or getting career-oriented tips and advice.

Being a job-oriented SNS, LinkedIn offers its users numerous benefits. For example, the job seekers can develop and post their resumes/CVs, make professional associates, publish skills that can be endorsed by other connected users; and obtain professional career guidance from the professionals. On the other hand, employers can advertise new job openings and find the right candidate they are looking for. Even, those users who are not looking for jobs right now can also join this site for getting career guidance tips, creating a professional network; and recommend a user for tentative hiring to the employers within their connections. Thus, it is not a surprising fact why the site is getting so popularity and why the number of users multiplying exponentially. This SNS's immense success and enormous number of active members may be linked to its core emphasis toward hiring opportunity (for job searchers) and selecting the best job applicants (for employers).

With over 58 million companies linked to this SNS and over 14m open opportunities, it can be anticipated that employers frequently utilize this SNS looking for the right candidates to fill their job positions (Khan, 2021). A report has revealed that in 2021, 122 million job candidates received interview calls while 35.5m of those interview call recipients later employed by the connected persons on LinkedIn (LinkedIn Press Center (n.d.). As reported by the LinkedIn Press Center (n.d.), three US job applicants each minute are being hired in USA through LinkedIn. In the last three-quarters of 2021, LinkedIn had over 3 million remote job postings, a 6 times increase than that of previous year (LinkedIn Press Center (n.d.). Moreover, another report revealed that candidates employed through LinkedIn profiles are 40% less likely to quit their organizations during the first six months on the job (Khan, 2021).

LinkedIn profiles provide several job-related details about candidates. In addition to their academic and professional credentials, users are allowed to describe their abilities, talents, volunteer works, computer/software skills and projects they have completed during the job search process (Roulin and Stronach, 2022; Roulin and Levashina, 2019; Shields and Levashina, 2016). Therefore, a user's LinkedIn profile can be substituted as an extension of his/her physical or online CV/resume (Zide et al., 2014; Kluemper, 2013). In contrast to traditional resumes, LinkedIn features such as connecting with certain other profiles, engaging with similar parties, posting thoughts or comments and following other connections or organizations are all examples of components that make LinkedIn different from other SNSs (Roulin and Stronach, 2022; Roulin and Levashina, 2019). The linked users in a user's network have the ability to support the published or claimed capabilities and such support may be seen on the users' profiles. The development of new capabilities that users themselves have not stated might also emerge through connections (Carr, 2016). In addition, LinkedIn members may contact their acquaintances for detailed written recommendations (Roulin and Stronach, 2022; Roulin and Levashina, 2019). In other words, LinkedIn profiles have many of the same features as traditional CVs/resumes and may also serve as a means through which references can be checked and reference papers can be obtained (Collmus et al., 2016).

Although, many academic research studies have examined using SM to find jobs (Hosain and Liu, 2020b, c); talent searching and hiring (Hosain et al., 2020a; Hosain and Liu, 2020a, d); and selection and pre-employment background checks (Hosain and Mustafi, 2023; Hosain, 2023; Hosain et al., 2020b, 2021; Hosain and Liu, 2019), quite a little similar research investigations have paid particular focus on LinkedIn. However, although a few studies (Roulin and Stronach, 2022; Damaschke, 2012; Khan, 2021; Hosain and Liu, 2020a, b; Roulin and Levashina, 2019; Kutlu et al., 2018; Chiang and Suen, 2015; Zide et al., 2014) have focused commonly at the function of LinkedIn profiles to determine the fundamentals that experts look for when recruiting a candidate; in particular, a very few papers (such as, Khan, 2021) focused on particular LinkedIn aspects like “LinkedIn Profile Richness (LPR)”, “LinkedIn Skill Recommendation (LSR)” and “Self-presentation on LinkedIn (SL)”. As far as the authors are informed, this is the first empirical study focusing on the role of LinkedIn features such as LSEs and LHRs on the Hiring Preferences (HPs) of Bangladeshi employers. Based on the study background and arguments made, the authors put forward the subsequent research question that this study tried to endeavor:

RQ.

What are the roles of LinkedIn-based Skill Endorsements and LinkedIn-based Hiring Recommendations on the Hiring Preferences of Bangladeshi employers?

Several sub-sections make up this paper. The first section of the paper presents the context and reasoning for the research. The second part of the paper offers a complete outline of the relevant current literature, formulates the research hypotheses and lays out its conceptual framework. The study methodology is described in section three while analysis and interpretation of the data have been presented in section four. The study outcomes and comparisons to earlier studies have been discussed in section five while the research implications (theoretical and practical) have been highlighted in section six. Section seven discusses the constraints of the research and the potential for further studies. The final section (8), wraps up the research.

2. Review of literature

2.1 LinkedIn and HP: earlier studies

LinkedIn is the largest professional and employment-related SNS which has over 850m regular memberships and 58m affiliated companies from over 200 countries and territories (Linkedin.com). The site is the largest (and maybe the only) professional SNS dedicated to finding work for its members, who range from fresh graduates to working professionals who are always on the lookout for qualified people to fill open jobs (Roulin and Stronach, 2022; Roulin and Levashina, 2019; Shields and Levashina, 2016). Since the members of each connected network engage with one another on a regular basis, it is easy for job candidates and recruiters to share information about job openings, internships, professional development opportunities and more (Hosain and Liu, 2020a, b).

A few studies focusing on LinkedIn and HP have been done in the recent 10–12 years. For instance, Peluchette and Karl (2010) found that students nearing graduation are increasingly conscious of the need for a distinct online identity, especially on LinkedIn, which is used for impression management and self-identification (Zide et al., 2014). Since LinkedIn was founded with the clear intention of connecting hiring professionals with qualified job seekers, recruiters are increasingly turning to job-focused SNSs like LinkedIn in their daily efforts to find and hire qualified candidates who meet their needs (Hosain et al., 2020a).

Numerous users' “Privacy” settings suggest that they do not want their SNS-based contents to be “Publicly disclosed”, despite knowing the fact that their accounts are connected within the web of accounts on various SNS. These users' materials may be labeled as “Private” (only content-holder user can view those materials). However, information placed on LinkedIn profiles are intended for the users to be seen by others within the same network, hence there is no such thing as “Private” or “Hidden” information on LinkedIn (Hosain et al., 2020a, b). LinkedIn, unlike Facebook and Twitter, provides prospective employees' academic achievements, professional experience, skill set and prior training. More so than on other SNSs, information uploaded on LinkedIn (such as abilities and experience) may be trusted as legitimate and true since it is vetted by the user's linked peers and prior work colleagues (Davison et al., 2011). Thus, more empirical research studies on the properties of LinkedIn-based HP are required.

2.2 LinkedIn-based skill endorsements (LSEs)

Applicants' LinkedIn profiles often include a lot of information that is relevant to the position they are applying for; including but not limited to, their schooling, awards, abilities, professional experience, projects completed and volunteer or community involvement (Shields and Levashina, 2016). It is a unique feature of a LinkedIn profile that a person's contacts inside the same network may verify, approve or suggest the user based on their claimed abilities. Additionally, the linked members can create skills about a user even if he or she has not posted any (Carr, 2016). LinkedIn's talent endorsement function stands out as an innovative and useful perk for the platform. According to Collmus et al. (2016), a user's endorsement count may be utilized as a strong signal of experience rank with proven talents, leading to higher ratings for that person.

According to Carr (2016), candidates are often more inclined to have the abilities they display on their LinkedIn profiles than on traditional printed resumes or CVs. This is due to greater “Warranting values,” which can be found on LinkedIn profiles (Roulin and Levashina, 2019). Colleagues and current/former employers can remark on whether they find the above talents to be a strength or weakness, respectively. As a result of this type of feedback, LinkedIn members feel more pressure to only list the abilities they actually have and can demonstrate competence in. Furthermore, endorsements allow the applicants' communication skills to shine through as a method for managing their online impressions by highlighting relevant talents (Roulin and Levashina, 2019; Shields and Levashina, 2016). What's more, as Roulin and Levashina (2019) point out, candidates with more networks and more recognized abilities tend to deliver more trustworthy and undistorted information. Employers are supposedly more likely to take seriously the claims of users with several skill endorsements. Therefore, a user's LSEs can be considered to be associated with higher positive ratings. The authors put forward the succeeding study hypothesis to be tested.

H1.

LSE has a positive role on HP.

2.3 LinkedIn-based hiring recommendations (LHRs)

Another unique characteristic of LinkedIn is that a user can request another user to write a recommendation letter. From the standpoint of the hiring organization, the quality of such LHRs is very significant (Roulin and Stronach, 2022). To explain in another form, a written recommendation based on LinkedIn profile connection is significant and creates higher chances of getting the expected job role. If the employers get a hiring recommendation from a former employer or a prominent individual, they positively associate it as the candidate's positive and reliable qualification (Roulin and Bangerter, 2013). However, on the other hand, if such a recommendation is not related to candidate's listed qualifications, the employer might have a negative impression on the candidate.

Vaughn et al. (2019) argue that an employer may value a job application more if it includes a reference letter from a decidedly competent user in a preferred skill domain and give them more confidence that the applicant is a good fit for the job. In their empirical study, Roulin and Levashina (2019) asserted that the likelihood of receiving a job offer or a promotion at the present company is highly correlated with hiring suggestions based on LinkedIn profiles. On the basis of prior study outcomes, the authors suggest that LHR also has a positive impact on HP and hence suggest testing the following research hypothesis:

H2.

LHR has a positive role on HP.

2.4 LinkedIn: the Bangladesh context

SNSs are growing quickly in South Asia, including Bangladesh. LinkedIn is becoming the top professional networking tool, yet Facebook is the most popular SNS in Bangladesh. LinkedIn is used by both the young and elderly Bangladeshis to find employment, acquire talent and create professional networks. Nearly 6 million LinkedIn users in Bangladesh were registered as of 2022, according to the Bangladesh Telecommunication Regulatory Commission (BTRC). Graduate students, employees, entrepreneurs and other professionals are all included in this number.

However, theoretical and/or empirical studies based on LinkedIn-based hiring in Bangladeshi job context are very rare except the study conducted by Khan (2021) and Hosain and Liu (2020a). Therefore, there is an urgent requirement for conducting further study investigations on this unexplored area of management.

2.5 Conceptual framework

This empirical study examines how LinkedIn attributes (LSE and LHR) affect Bangladeshi companies' recruiting decisions. Figure 1 serves as an illustration of the conceptual framework for this investigation.

3. Research design

3.1 Collection of data

For the purpose of this study, the researchers relied on primary data because the perception-based quantitative measurement cannot be well represented by secondary information. In order to obtain feedback from the respondents, the researchers made use of questionnaire items that were both thorough and organized. The respondents were purposely selected who work as the recruiters/hiring professionals. They have the expertise and are involved in SM-based (particularly, LinkedIn) selection. The respondents work as hiring professionals at a wide range of manufacturing and service firms in Bangladesh. Table A1 lists the respondents' sectors of occupations.

3.2 Measurement tool

The authors collected primary data using a 12-item survey (Table A2) with a “5-point Likert Scale” based on literature survey. Before sending out the questionnaires, the researchers briefed and trained the respondents on how to fill it out, reducing the likelihood of non-response bias, response bias and incompleteness.

3.3 Sampling method and sample pool

Due to the particular nature of this research, the researchers used purposive sampling strategy to recruit the study respondents. Purposive sampling is limited to precise categories of sample respondents who could deliver the desired data, either because they are the only right people to have that data or because they meet a number of other requirements (Sekaran and Bougie, 2010). If the researchers need the sample pool to meet a number of very specific requirements, then this sampling method is ideal (Blumberg et al., 2011). For this study, the researchers adopted “purposive sampling” to accumulate relevant information from the employers (Respondents) with experience in LinkedIn-based profile screening and hiring. In this study, first of all, the researchers selected those organizations that hire candidates based on their LinkedIn profile information as the supplement to regular CVs/resumes and other conventional sources of information. Secondly, from those selected organizations, the researchers selected those hiring professionals who are expert in hiring candidates' based on LinkedIn information.

The researchers primarily disbursed 500 research questionnaires to the institutional e-mails of the intended respondents. Following that, they got back 426 completed surveys (at a response rate of 85.2%). However, 38 of the received questionnaires were found to be incomplete, biased and/or unusually filled; and as a result, those questionnaires were removed from consideration. Finally, the total number of samples (n) was 388. Written permissions (from individual respondents) and institutional consent letters (from the organizations where they worked at the time of data collection) were obtained.

4. Analysis

4.1 Demographic information

As mentioned earlier, the respondents were chosen with care to form a cluster of representative sample who screen and recruit using LinkedIn. The demographic part of the survey instrument (Table 1) included gender, age, level of education, job experience; and LinkedIn-based monthly screening and hiring frequency.

Table 1 shows that the major portion of the respondents (88.66%) was male while half of them had age limit between 50–59 years. Almost all of them (97.42%) were Post-graduates with a few (2.58%) PhD degree holders. About two-thirds (61.60%) of those respondents had job experience of 30 years or more. Finally, 54.64% of the respondents screened and recruited candidates monthly 11–20 times based on their LinkedIn profiles, preceded by 1–10 times (25.52%).

4.2 Reliability and validity

To observe data reliability, the researchers used Composite Reliability (CR) and Average Variance Extracted (AVE) as presented in Table 3. The table indicates that all the CR values are higher than the allowable perimeter as recommended MacKinnon (2008), Hair et al. (1998), Fornell and Larcker (1981), and Henseler et al. (2009). They argued that CR value of 0.70 should be the permissible perimeter.

On the contrary, in order to test the discriminant validity, the authors opted for Fornell and Larcker's (1981) criterion that compares AVE value with equivalent correlation values of other constructs. According to those researchers, the square-root value of AVE should be higher than the equivalent correlation values of other constructs (Hair et al., 2014). The discriminant validity of every construct has been highlighted in Table 2.

4.3 Model evaluation

4.3.1 Measurement model (exploratory factor analysis)

As argued by Field (2000) and Hair et al. (1998), to assess Exploratory Factor Analysis (EFA), generally four recurrently used assumptions are followed.

1. Sampling adequacy (Kaisers–Mayesolkin) to be greater than 0.5

2. The minimum “Eigen Value” for a single factor

3. Being taken the sample into consideration, each item's factor loading should be 0.50 in order to be granted as the minimum limit to retain it with superior assurance; and

4. “Varimax Rotation” is required to be considered to be an established common notion that can simplify factor interpretations (Field, 2000).

Table 3 lists the outcomes of EFA. Hair et al. (2010) advocate that factor analysis can be carried out if “Kaiser-Meyer-Olkin (KMO) Test” and “Bartlett's Test of Sphericity” are critical. An index of Kaiser's measures of sampling adequacy (Overall MSA = 0.826) and Bartlett's Test of Sphericity χ² (p = 0.000) propose that the factor analysis was justified to conduct further analysis. After examining the EFA pattern matrix, the researchers identified that each and every item had an equivalent factor loading superior than 0.50 (Table 3).

According to the table, LSE may explain 16.60% of the overall variance having 4 items; LHR can explain 11.80% of the overall variance having 4 items, while the solo dependent variable (HP) can explain 15.30% of the overall variance having 4 items in this investigation. Moreover, the reliability values (Alpha) of all factors were well above 0.70 satisfying the essential criteria as advised by Nunnally and Bernstein (1994). Finally, all the 12 survey items were recognized to be suitable for further analysis.

The general result reveals that the EFA was appropriate. Subsequently, Maximum Likelihood (ML) and the Promax Method (PM) were deliberately used to identify the optimum model from those 12 items. Hair et al. (2010) recommend that an item's factor loading requires being more than 0.50 to be recognized as remarkably significant. Considering “Eigen Value” higher than 1, the researchers identified a three-factor model that can explain 43.70% of the overall variance of the dataset where those 12 items were grouped in to three dissimilar factors. The EFA outcome additionally identified 0.716 as the lowest and 0.916 as the highest factor loadings. The general outcome of the EFA, as identified, was satisfactory to conduct further analysis (Table 3).

4.3.2 Measurement model outcome (confirmatory factor analysis)

Confirmatory Factor Analysis (CFA) is a modern statistical procedure that can validate the factor structure of a set of experimental constructs. It allows the researchers to affirm whether the factors are associated with the corresponding variables. This model's related “Chi-Square” value was 3.211 which is lower than 5.0 as recommended by Marsh and Hocevar (1985) while additional fit indices as well pointed to a better fit to the measurement model. The “Goodness of Fit Index (GFI)” was 0.912 that is higher than the suggested limit of 0.90 (Joreskog and Sorbom, 1993). The authors summarized the outcomes of the measurement model and presented them in Table 4. As observed, all the fit indices exposed a superior match to the data.

In the present study, the “Adjusted Goodness of Fit Index (AGFI)” was found to be 0.872 that meets the approved value of (>0.85) and considered to be a better fit and supported by Anderson and Gerbing (1984). Moreover, the non-incremental fit index such as “Comparative Fit Index (CFI)” was 0.916 that has surpassed the recommended cut-off level of 0.90 (Bentler, 1990). In the CFA, the “Root Mean Residual (RMR)” value was found to be 0.076, less than 0.08, generally considered as accepted (Hu and Bentler, 1999). The “Root Means Square Error of Approximation (RMSEA)” was 0.077 that is also less than the suggested good fit (0.08) (Browne and Cudeck, 1993). Finally, the “Standardized Means Square Residual (SRMR)” was identified as 0.076; less than 0.08 and tolerable as recommended by Browne and Cudeck (1993).

4.4 Assessment of multicollinearity

In order to discover the inter-factor multicollinearity, the researchers considered “Variance Inflation Factors (VIFs)”. VIFs can vary from 1 to 10 and beyond. The VIF numerals can clarify the percentage of variance that is inflated for each co-efficient. A commonly established principle regarding the interpretation of VIF is that 1 or less represents no correlation, 1 and 5 represents moderate correlation; and a VIF value higher than 5 indicates extreme correlation (Hair et al., 1998).

To observe the multicollinearity level, the researchers calculated the VIFs (Table 5). It was discovered that the maximum VIF value was 2.101 as presented in Table 5 which is inside the satisfactory boundary as recommended by Hair et al. (1998).

4.5 Common method bias (CMB) test

When the superior fraction of the variance can be explained by a single factor, Harman's Single Factor Model can be used (Rehman et al., 2022; Podsakoff and Organ, 1986). If a single factor does not explain the greater portion of variance, it can be presumed that no CMB has been occurred (Aguirre-Urreta and Hu, 2019).

As observed from Table 6, it can be concluded that a particular factor in this model could explain only 16.546% of variance which indicates that CMB did not occur in this analysis.

4.6 Structural model

The researchers used CB-SEM technique to reveal the influence of two LinkedIn-based factors (LSE and LHR) on HP. The statistical outcomes have been presented in Table 7 with the structural parameter estimates and the hypotheses testing outcomes.

The results discovered that LSE has an insignificant role (β = 0.18; CR = 2.262; p > 0.05) on HP rejecting the first hypothesis (H1) while LHR has a significant positive role (β = 0.22; CR = 2.316; p < 0.05) on HP supporting the second hypothesis (H2). The two independent factors can effectively justify 47.2% of the overall variance in the dependent variable according to this CB-SEM model that have been indicated in Table 7 and Figure 2.

5. Discussion

The statistical outcomes revealed that LSE has an insignificant role (β = 0.18, CR = 2.262; p > 0.05) on the recruiting decisions of employers in Bangladesh. A result like that runs contrary to what the scholars previously reported. However, regarding the second factor, LHR, the researchers found that it has a noteworthy positive role (β = 0.22, CR = 2.316 and p < 0.05) on HP. The outcome is supportive to the earlier outcomes reported by Roulin and Stronach (2022), Khan (2021), Roulin and Levashina (2019), Vaughn et al. (2019), and Roulin and Bangerter (2013).

Therefore, one of the hypotheses (H2) proved valid while the other one (H1) proved invalid based on the perceptions of employers in the Bangladeshi job context. However, according to all of the respondents, LinkedIn is a powerful SNS that is regularly used for choosing competent talents. The findings of this empirical study can represent LinkedIn's effectiveness as a legitimate recruiting tool in an emerging country context because of the country's high population density and large number of young people searching for employment (Khan, 2021).

6. Implications for theory and practice

6.1 Theoretical implications

Even though a considerable number of theoretical and empirical research papers have been published on the influence of SNSs on various HRM practices (Hosain, 2023; Roulin and Stronach, 2022; Hosain et al., 2020a, b; Khan, 2021; Hosain and Liu, 2020a, b, c, d), only a few of those publications have examined the roles of LinkedIn-based characteristics on employability or hiring preferences. Particularly, employers' perceptions concerning LinkedIn-based job features as the predictors of hiring preferences have been rarely identified in the academia so far despite of its growing popularity among both the job-seekers and the recruiters. As one of the initial attempts to identify the impact of LinkedIn-based job features on the hiring preferences of employers in an emerging economy, the authors strongly believe that this study can influence and motivate the upcoming researchers in conducting their empirical studies in a developing country context.

LinkedIn is gaining significant momentum on the job sectors not only in Bangladesh but around the world. However, considering such a significant impact of this SNS among the employers and jobseekers, the academic attention on LinkedIn-based employability is comparatively less focused. In this empirical investigation, the authors tried to highlight that unattended area to some extent. Furthermore, a substantial number of these published works focused only on western nation settings. As one of the largest (in terms of population size) and labor abundant countries in the world, Bangladesh lacks empirical evidence about LinkedIn's impact on employment preference. This article is one of the earliest efforts to address this theoretical gap. As one of the pioneering investigations concerning the role of LinkedIn-based job features on hiring preference, the authors strongly believe that this wider empirical study will not only fill the acute lack of LinkedIn-based scientific papers but also reduce the gap between the developed and developing county contexts regarding this issue.

6.2 Practical implications

This experiential study is anticipated to assist and guide the administrative policymakers (such as, HR professionals and/or hiring staffs) to devise and execute a suitable, concrete and well-timed organizational SNS-based recruitment/selection framework and associated policies/procedures. SNSs with a focus on the workplace, like LinkedIn, are becoming increasingly mainstream online tools that, when used with caution and in accordance with established protocols may save costs and free up important time. Most importantly, this empirical report could be a critical assistance for designing and enforcing a suitable SM strategy, resulting in the reasonable and competent use of social networking information for the hiring process. Thus, our experimental paper is anticipated to reduce the gap between the theory and practice to some extent.

From this study outcomes, the hiring professionals of the developing countries can have some basic guidelines and tips on how to effectively utilize LinkedIn-based characteristics to find a competent job candidate. More importantly, the authors believe that this study can aid in formulating and executing effective social media policies and practices that can make the hiring decisions more efficient.

7. Study limitations and further research scope

Although being a pioneering study in Bangladeshi context concerning LinkedIn-based features and hiring preferences, this paper has some obvious limitations that the authors would like to admit. First of all, a certain cultural backdrop was taken into consideration where the research was conducted. Possibly more insightful cross-cultural and comparative research studies would have provided more representative outcomes. As independent variables, the authors considered only two constructs. The results could have been more representative if additional factors had been used. Finally, it is possible to further expand the sample size.

The authors anticipate that more studies will examine how LinkedIn-based job attributes affect recruiting. The authors believe that while performing their empirical study on how LinkedIn affects the total employment process; the upcoming researchers will take these limitations into account as the research gaps.

8. Conclusion

Based on the research findings of this study, it is clear that a LinkedIn profile's components can positively impact the recruitment decisions of Bangladeshi companies. This indicates that the website has a strong image and is acceptable to both employers and employees in the context of the Bangladeshi job market. Indeed, as the main employment-oriented SNS, LinkedIn is popular not just with employers, job prospects and passive job searchers; but also as a medium for sharing professional experience and building networks. Such popularity is merely expected to grow in the upcoming years due to many distinct employment features that may ease the demands of various linked parties about their individual requirements.

SNSs are unquestionably modern science and information technology's gifts. We can't entirely escape them even if we wanted to, but we can make the best of them. Based on user requirements and preferences, it is expected that more and more job-related attributes will be added to this site. Employers and job seekers may both benefit from using a job-focused SNS like LinkedIn if they pay close attention and care. LinkedIn may be the easiest option for passive job searchers who cannot submit job applications by the correct channels owing to a lack of time and expertise. Furthermore, the site offers much job-related information that benefits professional development and career advancement.

LinkedIn is clearly the most popular job-oriented SNS; and it is expected to become an indispensable employment feed soon. Utilizing simply LinkedIn profiles for hiring, however, as a replacement for the traditional CV/resume may take a lot of work, at least at the initial level. Moreover, using LinkedIn as the only hiring method may result in legal concerns. Likewise, LinkedIn accounts may include several non-job-connected contents/information (photos and other personal data) that can generate decision making biases in the employment process. Nevertheless, using this site as a hiring tool is an enormously rising practice in the corporate sector, the issue of using it as a recruitment and selection tool is severely under-researched in the academia. Thus, the authors expect that this empirical paper will motivate the researchers for conducting further empirical studies focusing on research limitations that have been mentioned in the previous section.

Figures

Figure 1

Conceptual framework

Figure 2

Theoretical framework with hypotheses testing outcomes

Table 1

Respondents' demography

Demographic aspects	Category	Absolute number	Proportion
Gender	Male	344	88.66
	Female	44	11.34
	Total (n)	388	100
Age range (year)	30–39	39	10.05
	40–49	152	39.18
	50–59	197	50.77
	Total (n)	388	100
Educational level	Post-graduation (Master)	378	97.42
	PhD	10	2.58
	Total (n)	388	100
Experience (year)	10–19	37	9.54
	20–29	112	28.86
	30+	239	61.60
	Total (n)	388	100
LinkedIn-based screening and hiring frequency (Monthly)	1–10	99	25.52
	11–20	212	54.64
	21–30	77	19.84
	Total (n)	388	100

Source(s): Survey instrument

Table 2

Discriminant validity

Variable/Construct	HP	LSE	LHR
HP	0.89
LSE	0.09**	0.82
LHR	0.201*	0.238***	0.77

Note(s): p < 0.100, *p < 0.050, **p < 0.010, ***p < 0.001

Source(s): SPSS 25

Table 3

EFA outcomes

Factors	Eigenvalue	Percent of variance	Cumulative variance	Survey items	Loading	CR	AVE	Alpha
LSE	3.83	16.60	16.60	LSE1: One distinctive feature of LinkedIn is the capability of skill endorsement	0.916	0.92	0.81	0.901
				LSE2: Since colleagues or past employers have verified the candidates' claimed skills, their capabilities are real	0.872
				LSE3: LinkedIn-based endorsed skills are particularly important for getting a job offer	0.812
				LSE4: If required, I match the endorsed skills and the skills listed on a candidate's CV/resume	0.823
LHR	3.73	11.80	28.40	LHR1: LinkedIn has a special feature that allows for hiring recommendations	0.813	0.84	0.66	0.841
				LHR2: Written recommendations particularly by the previous employer play a vital role in getting a job offer	0.759
				LHR3: We prioritize those job candidates who can bring written recommendation letters	0.772
				LHR4: LinkedIn-based hiring recommendation is an effective outcome of professional networking	0.716
HP	4.85	15.30	43.70	HP1: LinkedIn is a unique site that is a good medium for professional networking	0.871	0.91	0.82	0.816
				HP2: Those who are active in LinkedIn have a better chance of being hired	0.742
				HP3: The tiresome paperwork involved in identifying qualified job candidates has been minimized by LinkedIn	0.724
				HP4: Employers are more frequently relying on LinkedIn to locate and hire qualified candidates	0.814

Note(s): KMO = 0.826, Degree of Freedom = 251, Significance = 0.000

Source(s): SPSS 25

Table 4

Model fit indices and their recommended limits

Fit indices	Present value	Acceptance level	Reference
Chi-square/df	3.211	<5.0	Marsh and Hocevar (1985)
CFI	0.916	>0.90	Bentler (1990)
RMR	0.076	<0.08	Hu and Bentler (1999)
GFI	0.912	>0.90	Joreskog and Sorbom (1993)
AGFI	0.872	>0.85	Anderson and Gerbing (1984)
RMSEA	0.077	<0.08	Browne and Cudeck (1993)
SRMR	0.076	<0.08

Source(s): AMOS 25 and literature review

Table 5

VIFs and multicollinearity tolerance

Tolerance	0.705	0.901	0.703	0.694	0.472	0.561
VIF	1.447	1.855	1.326	1.221	2.101	1.743

Source(s): SPSS 25

Table 6

CMB test

Total variance explained
Component	Initial eigen values			Extraction sums of squared loadings
	Overall	Percentage of variance	Cumulative percentage	Overall	Percentage of variance	Cumulative percentage
1	4.202	24.007	24.007	4.741	16.546	16.546

Source(s): SPSS 25

Table 7

Regression weights (group number 1 – default model)

Path	Hypothesis	Estimate	Standard error	Critical ratio	p value	Label
HP←LSE	H1	0.18	0.075	2.262	0.051	Not Supported
HP←LHR	H2	0.22	0.064	2.316	0.006	Supported
HP		0.472

Source(s): AMOS 25

Table A1

Sector-wise distribution of the respondents

List of industrial sector	Number of firms	Number of respondents
Manufacturing	11	41
Telecommunication	10	39
Financial institutions	11	55
Academic institution	5	36
IT firm	8	47
Transportation	2	11
Freight/Shipping	3	18
Postal service	1	13
Media (print and electronic)	6	29
Food and beverage	12	44
Hotel and tourism	9	55
Total	78	388

Source(s): Survey instrument

Table A2

Questionnaire items with literature sources

Factor name	Items	Literature sources
LSE	LSE1: One distinctive feature of LinkedIn is the capability of skill endorsement	Khan (2021), Roulin and Levashina (2019), Shields and Levashina (2016), Collmus et al. (2016), Carr (2016)
	LSE2: Since colleagues or past employers have verified the candidates' claimed skills, their capabilities are real
	LSE3: LinkedIn-based endorsed skills are particularly important for getting a job offer
	LSE4: If required, I match the endorsed skills and the skills listed on a candidate's CV/resume
LHR	LHR1: The option of hiring recommendation is a unique characteristic of LinkedIn	Roulin and Stronach (2022), Roulin and Levashina (2019), Vaughn et al. (2019), Roulin and Bangerter (2013)
	LHR2: Written recommendations particularly by the previous employer play a vital role in getting a job offer
	LHR3: We prioritize those job candidates who can bring written recommendation letters
	LHR4: LinkedIn-based hiring recommendation is an effective outcome of professional networking
HP	HP1: LinkedIn is a unique site that is good medium for professional networking	Hosain and Liu (2020a), Hosain et al. (2020a), Zide et al. (2014)
	HP2: Those who are active in LinkedIn have a better chance of being hired
	HP3: The tiresome paperwork involved in identifying qualified job candidates has been minimized by LinkedIn
	HP4: Employers are more frequently relying on LinkedIn to locate and hire qualified candidates

Source(s): Literature survey

Source of funding: No source of funding to be reported.

Conflict of interests: The authors declare no conflict of interests.