Influence of Recruitment Practice on Lecturers’ Performance in Kenyan Public Universities

Public universities are key players in pursuit of Kenya Vision 2030’s mandate of providing globally competitive quality education, training and research to the citizenry. However, higher education stakeholders including the World Bank, Kenya’s Commission for University Education, employers and scholars have expressed misgivings regarding quality and relevance of these institution’s academic programmes. Declining quality is attributed to increased workload, reduced rigour in recruitment and promotion criteria, failure to attract and retain quality faculty, and alleged lecturers’ incompetence. Empirical studies relating recruitment practice (RP) to employee performance (EP) ignore both descriptors of RP like: examining vacancy; locating, making contact and attracting candidates as well as lecturers’ performance measures such as: quality of teaching and learning; research and publication; administration and other responsibilities; and community engagement and other contributions. The purpose of this study was to establish influence of recruitment on lecturers’ performance in Kenyan public universities. It was guided by Simon’s decision making theory and reinforced by Barney’s resource based view; Adam Smith’s human capital theory; and Herzberg’s two-factor theory. It adopted descriptive cross-sectional survey research design with a target population of 1,653 lecturers in 14 chartered public universities established between 2009 and 2019. Multi-stage purposeful sampling technique was used to finally take a census of 158 academic heads of departments (lecturers) from 13 out of the 14 universities, having used 20 in piloting. Public universities were targeted since they account for 72% of the teaching total university teaching faculty. Semi-structured questionnaires were used to collect primary data while secondary data were obtained from relevant secondary sources. Cronbach’s alpha coefficient was used to examine the internal consistency at 0.796 (SD= 0.067). Whereas content validity was ascertained by reviewing literature within the domain of study concepts and corroborated by a panel of experts, construct validity was measured using Pearson’s r. Data were analyzed using Pearson’s r and regression analysis. The correlation between recruitment practice and lecturers’ performance was established to be positive and significant ( r =.324, p =.000). Recruitment practice accounted for 11.6% (Adjusted R 2 =.116) of the variation in the lecturers’ performance in Kenyan public universities. The findings are useful to the management of public universities and HRM practitioners in formulating and implementing recruitment strategies; researchers as well as organizational behaviourists and theorists.


Introduction
Higher education seeks to improve the social life of a nation (Teir and Zhang, 2016) given its importance in the promotion of social equity and economic security (Hudson, 2016;Malee-Baset, 2015). Higher education facilitates economic transformation by supporting lower levels of education, preparing professionals and skilled labour, as well as serving as an incubator for research (World Bank, 2017). It is against this backdrop that Kenya Vision 2030 was conceived to among others provide globally competitive quality education, training and research to her citizens for development and enhanced individual wellbeing (Republic of Kenya, 2007). Higher education comprises a range of higher learning institutions including the university (Alemu, 2018). The role of universities in preparing human resources for sustainable development of nations is monumental (Kara, Tanui and Kalai, 2020). Ironically, the quality and relevance of higher education, value for money and resource allocation in the sector have raised a lot of concerns among stakeholders (World Bank, 2017;Johnson, Kirimi and Gathara, 2018;Ball, 2012;Adeyemi, 2017).
The resource based view was propounded by Barney (1991). It perceives human resources as important factors in attaining competitive advantage, and realizing organizational goals and targets (Barney, 2001). Strategic capability of an organization, according to the theory, is premised in its resource capability in form of her employees (Armstrong, 2009). HRM adds value through the strategic development of the rare, hard to imitate and hard to substitute human resources (Barney, 1991). In respect of this study, the theory draws its relevance from its focus on the quality of human resources available to an organization and their ability to adapt more quickly relative to their competitors (Katua and Gachunga, 2014) hence optimum performance. Barney (2001) encourages firms to constantly evaluate their workforce to ensure that they have the right people with the right skills in the right places to assure sustained competitive advantage. In the event of any shortfalls, firms are advised to employ appropriate recruitment and selection criteria (Ekwaoba, Ikeije and Ufoma, 2015) or, in the alternative, initiate training and development programmes for the existing employees so as to enhance their knowledge, skills and abilities.
The stance of human capital theory that was propounded by Adam Smith (Terence, 1976) is that people are fixed capitals, just like machines owing to the fact that their skills and abilities have costs and yield profits (Teixeira, 2002). It is premised on the assertion that people and their collective skills, abilities and experience, coupled with their ability to deploy these in the interest of the firm contribute to its competitive advantage and success (Armstrong, 2006). Boxall (1996) views this situation as one that confers "human capital advantage". According to the theory, sustainable competitive advantage is attained when the firm has a human resource pool that cannot be imitated or substituted by its rivals. Firms which recruit and retain exceptional individuals can create competitive advantage for themselves (Boxall, 1998) in addition to generally enhancing their performance.

Recruitment
According to Nzuve (2010), recruitment is a fundamental function in the management of human resources in large organizations. It is the first part of the process of filling a vacancy which includes: the examination of a vacancy, locating candidates, making contact and attracting applications from suitable candidates (Graham and Bennett, 1998). It is aimed at finding and engaging the people needed by an organization (Armstrong, 2009). DeCenzo and Robbins (1998) define it as the discovering of potential candidates for actual or anticipated organizational vacancies. It should attract a large number of qualified candidates to choose from, and provide information that enables unqualified applicants to self-select themselves out of the job candidacy. The process of recruitment is only complete in as far as it attracts qualified job applicants (Nzuve, 2010). Recruitment can be internal or external. Internal recruitment implies looking inside the organization for existing qualified employees who might be promoted to higher positions while external recruitment refers to looking outside the organization for prospective employees (Graham and Bennett, 1998;Armstrong, 2009;Nzuve, 2010;DeCenzo and Robbins, 1998;DeNisi and Griffins, 1998). For recruitment to achieve its objective, DeCenzo and Robbins (1998) suggest that recruiting sources should reflect the type of position to be filled. While both internal and external recruitment have a range of merits and demerits (Armstrong, 2006;Armstrong, 2009;Dessler, 1998;DeNisi and Griffin, 1998;DeCenzo and Robbins, 1998;Beardwell and Claydon, 2007;Graham and Bennett, 1998;Torrington et al. 2008;Ivancevich, 2010), Armstrong (2009) opines that priority should be given to the internal candidates. As such, external recruitment is only recommended when an organization is unable to get a suitable candidate from within. DeNisi and Griffin (1998) elucidate three methods of internal recruitment: job posting, supervisor recommendations and union halls. Dessler (2008) expands the list by adding personnel records and skill banks. The most common methods of external recruitment are: online or e-recruitment; advertising; employment agencies; referrals, recommendations and walk-ins; educational institutions; temporary help services and professional organizations (Armstrong, 2009;Dessler, 1998;DeNisi and Griffin, 1998;DeCenzo and Robbins, 1998;Beardwell and Claydon, 2007;Graham and Bennett, 1998;Torrington et al. 2008). They note that each of the methods has its merits and demerits.

Lecturers' Performance
Performance is a multi-dimensional construct (Dugguh & Ayaga, 2014). Performance in organizational context is broad and has been a subject of study among social scientists from a wide range of disciplines, and can be used synonymously with productivity, efficiency, effectiveness and, more recently, competitiveness (Cooke, 2001). Performance incorporates behaviour and results (Brumbach, 1998;Dugguh & Ayaga, 2014). Behaviour emanates from performer and transforms performance from abstraction to action. Employee performance implies the jobrelated activities expected of employees and how they are executed (Dugguh and Ayaga, 2014). Lecturers are employees of universities charged with transforming developing and disseminating knowledge through education, research and community service. Consequently, their performance is primary to the success of a university. Lecturers' performance can, therefore, be defined as the job related behaviours and results expected of lecturers European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021 (Brumbach, 1998;Dugguh and Ayaga, 2014).
There are divergent indicators of lecturers' performance. A measure is a specific quantitative or qualitative representation of capacity, process or outcome deemed relevant to the assessment of performance. Lecturers' performance can be measured by subject knowledge, testing or assessment procedure, student-teacher relations, organizational skills, communication skills, subject relevance and utility of assignments (Molefe, 2010). Sánchez-Barrioluengo (2013) in World Bank (2017) explains that the higher education community generally identifies three distinct but interrelated missions: teaching and learning, research and community engagement against which the performance of lecturers could be measured. Abba and Mugizi (2018) summarize the performance of academic staff as being indicated by: teaching, research and publication, innovation and community service while Commission for University Education (2014a and 2014b) wrap lecturers' performance around four thematic areas: quality of teaching and learning; research and publication; administration and responsibilities and community engagement and other contributions. As a policy, the commission embedded the appointment and promotion of academic staff in public universities in the four areas through a document entitled "Harmonized Criteria and Guidelines for Appointment and Promotion of Academic Staff in Universities in Kenya". While the document was challenged in a court of law by University Academic Staff Union, the contention was not on the performance areas. Universities have, therefore, largely retained the performance areas with some variations on weighting. This study therefore, adapted the Commission for University Education's indicators, and blended them with relevant dimensions of measures proposed by other researchers in Molefe (2010).
Quality of teaching and learning is measured by the following: student evaluation of the instructor and course; the lecturer's notes; ability to advise and mentor students; training in higher education teaching and post graduate supervision (Commission for University Education of Kenya, 2014a and2014b). Whereas student evaluation may court controversy, Balam and Shannon (2010); and Bedggood and Donovan (2012) submit that there is sufficient literature to conclude that they are generally consistent and valid (Arnold, 2008 andLiu, 2012). Related to these is a lecturer's knowledge-base in the subject as proposed by Analoui (2007) who views it as being fundamental to the creation and enhancement of the student's opportunity to learn well. Knowledge base encompasses declarative knowledge of facts and concepts, the procedural knowledge of what to do, and the motivation includes the effort and persistence to excel (Aguinis, 2009). Ganyaupfu (2013) adds the teaching skills of a lecturer to this list. He explains that a lecturer's teaching skills is based on their ability to comprehend and transform concepts to be imparted to learners. Shulman (1992) views comprehension as a key element of lecturers' competence. Sinclear and Johnson (2000) assert that thorough knowledge of the subject material is essential to accurate instruction and clear communication and content to students. Arreola (2000) opines that the quality of teaching and learning should also include expertise in the content area.
In addition to content knowledge, ability to organize, integrate, adjust and adapt this content in ways that make it accessible and thought provoking to the learner are the other parameters of quality of teaching and learning (Gill and Johnson, 1997;White, 2008). Testing or assessment procedure which entails designing, developing and implementing tools and procedures for assessing students' learning outcomes and is part of the instructional design (Molefe, 2010) also contributes to the quality of teaching. Arreola (2000) identifies requisite skills under this dimension as: designing tests; preparing the learning objectives; developing syllabi; preparing handouts and other supporting materials; properly using media and other forms of instructional technology and organizing lectures and presentations for maximal instruction. Hill, Lomus and MacGregor (2003) add feedback to students during the sessions and assignments to this list. Molefe (2010) further points out the importance of student-teacher relations in teaching and learning. He relates it to the creation and maintenance of student-centred environment that maintains and sustains learning and development. According to him, it is a dimension that is integral to high learner-performance. Arreola (2000) asserts that a teacher who can develop relationships that foster and encourage students will enhance learning. Encouragement of active participation in the classroom creates a supportive environment where questions and class discussions are promoted, which imbues the lecturer with enthusiasm for the subject and facilitates opportunities for generating regular informal feedback on students as well as deeper understanding of the subject matter (Sinclair and Johnson, 2000). Sinclair and Johnson (2000) opine that organizing skills is another dimension that influences overall student experiences, as well as the quality of teaching. Included herein is also the bureaucratic skills utilised for operating and managing a course including, but not limited to, timely grading of examinations, maintaining published office hours, arranging for and coordinating guest lecturers and generally making arrangements for facilities and European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021 resources required in the teaching of the course. In support of this stance is Arreola (2000) who asserts that excellent teachers do their work in a well-prepared and well-organized manner by arranging their activities in a way that allows them time to engage in corporate citizenship and community outreach. It is worth noting that corporate citizenship and community outreach are other measures of lecturer performance according to the Commission for University Education (2014a and 2014b). Molefe (2010) further identifies communication as an important aspect of structural delivery skills that enhances teaching and learning. Clarity in exposition, demonstrated enthusiasm, ability to motivate, ability to capture and hold the interest and attention of learners and create an overall learning environment appropriate to the content being taught are all included in communications skill dimension (Arreola, 2000). According to Hill et al., 2003 andWhite, 2008, subject relevance is another factor that concerns learning and teaching quality. It relates to the appropriateness of the content provided during the lesson and the way in which it is presented to the learners. They suggest that it should entail accuracy of the facts encapsulated in the course content. Relevant assessment instruments used in the course should add to the design of architecture within the frame of reference of the course materials and the real world associated with the subject. Questions should be set at the appropriate level and graded according to the learning outcomes of the module. The text books and reference materials recommended by the lecturers, as well as the appropriate use of methods and techniques used in the subject are also of vital importance. The course being offered should also be of value to the workplace.
Towards enhancing the quality of teaching and learning, White (2008) opines that the assignments given to students should be meaningful and ones that enhance their learning and developmental needs. The assignments should be within the frame of reference of the course materials and the real world associated with the subject and socioeconomic life within which a student lives. To further enhance utility, Layman, Williams and Staten (2007) suggest that the assignment should reflect its learning objectives and make it interesting and challenging to the student.
Research and publication as an indicator of lecturers' performance generally implies the ability to advance scholarship and generate research as advanced by Gill and Johnson (1997) and White (2008). Implicit in the generation of research is the advancement of knowledge through discovery, integration, dissemination and application of knowledge. Commission for University Education (2014b) suggests that it entails: publication of scholarly books and chapters in books; refereed journals; refereed learning modules; reviewed conference papers; non-reviewed conference papers; short communication in a refereed / scholarly journals; book reviews published in refereed journals; editorship of books and conference proceedings; scholarly presentations at conferences, workshops or seminars; refereed exhibitions and performances; consultancy and project reports; and patented invention and innovation. While these measures advance a quantitative component (i.e. numbers of), Harvey et al.  (2003) in Cadez, Dimovski and Groff (2017) suggest that the quality aspect be interrogated via the impact and quantity of papers published in an exclusive set of high quality journals or by the number of citations.
Administration and responsibilities as a measure of lecturers' performance encapsulates recognized university administrative positions and any other responsibilities that may be assigned to a lecturer. Community engagement and other contributions, on the other hand, is rated by the ability of a lecturer to attract research and development funding; engagement in community service and outreach, professional affiliations and portfolios, recognition, awards and honours amongst others as determined by individual universities (Commission for University Education, 2014b).

Recruitment Practice and Lecturers' Performance
Reviewed studies relate recruitment directly or indirectly to either employee or organizational performance. Saifalislam, Osman and AlQudah (2014) not only relate recruitment to organizational performance, which is a function of many other variables, but fail to specify the indicators of recruitment. Suwarto (2018) whose interest was on ethical issues that affect recruitment basically identified unethical practices that need not characterize recruitment such as sexual harassment, bribery, nepotism and discrimination; Oaya, Ogbu and Remilekun (2017) focused on recruitment strategy that they operationalized to comprise only three methods of recruitment: employee referrals, employment agencies and host community member recruitment; European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021 Mokaya, Mukhweso and Njuguna (2013) considered recruitment sources, recruitment policies and recruitment message as the descriptors of recruitment practice. Both Suntanto and Kurniawan (2016) and Ombui, Mukulu and Waititu (2014) are not specific on the elements of recruitment they studied despite its varied components. Therefore, none of them studied the facets of recruitment such as examining vacancy; locating, making contact and attracting candidates (Graham and Bennett, 1998).
Revealing a positive significant relationship between the various facets of recruitment and employee performance are Rahmany (2018) (2013) found a positive significant, but weaker relationship. Oaya, Ogbu and Remilekun (2017) discovered a significant positive relationship between employee referral recruitment and employment agency recruitment (facets of recruitment strategy), and employee commitment and productivity (facets of employee performance) respectively. However, host community member recruitment revealed no significant relationship with employee efficiency (facet of employee performance).  established that recruitment had no effect on employee performance. The results were, therefore, mixed.

Research Design
Research design refers to the arrangements of the conditions for collection and analysis of data in a manner that aims to combine relevance to the research purpose with economy in procedure (Kothari 2004). The study adopted descriptive cross-sectional survey research design in data collection and analysis. A survey research design is scientific method of inquiry in which the researcher selects a sample of respondents and administers a standardized research instrument on them (Rubin and Babbie, 2001) define. Descriptive research studies describe the characteristics of a particular individual, or of a group and facilitate predictions, with narration of facts and characteristics concerning individuals, groups or situations (Kothari and Garg, 2014). Creswell (2009) observes that surveys as forms of quantitative strategies can provide quantitative or numeric description of trends, attitudes, or opinions of a population by studying a sample of that population. The study used academic heads of departments as the unit of analysis. Cross-sectional survey was preferred because it facilitated the collection of data from various cases at the same time hence a variety of views over the same issue captured within a short time. The correlational approach assisted in establishing whether there existed a significant association between the independent and the dependent variables (Mugenda, 2008;Kothari and Garg, 2014). Creswell (2009) opines that questionnaires can be used to collect data in cross-sectional study.

Study Area
The study focused on 14 chartered public universities that were established between 2009 and 2019, but spread across the Republic of Kenya. Kenya lies between Latitude: 0° 10' 36.73" N and Longitude: 37° 54' 29.98" E, and on the equator with the Indian Ocean to the South-East, Tanzania to the South, Uganda to the West, South Sudan to the North-West, Ethiopia to the North and Somalia to the North-East. The country was targeted due to the misgivings that Higher Education stakeholders have expressed regarding the quality of education in her universities (Mwiria and Nyukuri, 1994;Ngolovoi, 2006;Wangenge-Ouma, 2007;Gudo, Olel and Oanda (2011); Kaburu (2014) and Ayiro (2015).

Target Population
Target population is the population of interest in a given study. It is the theoretical population (Umar, 2018). According to Bartlett et al. (2001), target population is the group of individuals or participants with specific attributes of interest and relevance. It is, therefore, the part of general population that is left after refining and eliminating all individuals whose involvement in the study would breach the research goal, assumptions or context. The target population of the study comprised all the 1,653 lecturers in the fourteen (14) chartered public universities established between 2009 and 2019. Any lecturer outside this specification was, therefore, excluded.
Accessible population is arrived at after eliminating all individuals who may not participate or who cannot be accessed in the course of the study (Bartlett et al., 2001). It constitutes the final group from which data is collected in whole or as a sample. It, therefore, represents the sampling frame. Accordingly, the accessible population of the

Sample Design and Size
Sample design is a definite plan for obtaining a sample from a given population. It refers to the technique or procedure that the researcher adopts in selecting items for a sample. It may lay down the number of items to be included in the sample (Kothari, 2004). The study adopted multistage purposeful sampling technique (Table 1) in which the study sample was selected in two stages purposively (Omona, 2013;Onwuegbizie and Leech, 2007). In the first stage, the study purposively focused on 1,653 lecturers in the 14 chartered public universities established between 2009 and 2019 (its target population) , from the general / entire population of 8,737 lecturers in the 31 chartered public universities in Kenya as at 20 th October, 2019. The universities 14 universities were selected on the basis of the staffing challenges they are faced with (Ikama, 2010: Kagondu andMarwa, 2017) and the pressure exerted on them to employ (Mukhwana et al., 2016).
Judgment sampling was then purposively employed, in the second stage, to take a census of 178 academic heads of departments (AHoDs) in the 14 chartered public universities since they were considered information rich. The AHoDs normally participate in making recruitment decisions. They are also the immediate supervisors of the lecturers in their respective departments hence better placed to monitor and assess their performance. Their choice was further qualified by the fact that they are representative of all the categories of lecturers considering that they are drawn from disparate ages, gender, academic positions or ranks and qualifications across the universities apart from being lecturers themselves. The deans and directors of schools and institutes that were not established as departments under existing schools or faculties were considered the AHoDs. The census of 158 was adopted due to the limited number that remained after conducting a pilot study on 20 AHoDs from Jaramogi Oginga Odinga University of Science and Technology (JOOUST). JOOUST was selected using simple random sampling technique. Monette, Sullivan and Dejong (2002) suggest that a sample of 20 respondents is sufficient for pilot testing in a survey.

Data Type and Sources
Both primary and secondary data were collected. Primary data was drawn from the respective AHoDs using semistructured questionnaires while secondary data extracted from journal articles, human resource records, audit reports, relevant books and institutional websites. Cooper and Schindler (2008) applaud primary data for their proximity to the truth and control over error.

Data Collection Procedures
Prior to the administration of the data collection instrument, the researcher first obtained authorization to collect data from the School of Graduate Studies, Maseno University; Maseno University Ethical Review Committee; and research permit from National Commission for Science, Technology and Innovation. Permission to collect data was then sought and obtained from the administration of the respective universities that comprised the study. This was followed by an informal familiarization visit to the targeted universities to create rapport and locate the relevant offices, and respondents. It is at this point that the researcher obtained tentative dates and time for the European Journal of Business and Management www.iiste.org ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.13, No.20, 2021 151 administration of the questionnaires at the convenience of the respondents.
The research questionnaire had hitherto been given to a panel of seven experts, in consultation with the researcher's supervisors, for review. Having incorporated their suggestions which included summarizing the items that were considered wordy, use of common words, reconstruction of poorly constructed item and blending of negative and positive statements in the questionnaire items, it was then piloted on twenty (20) AHoDs drawn from Jaramogi Oginga Odinga University of Science and Technology and their suggestions incorporated too. Monette, Sullivan and Dejong (2002) suggest that a sample of 20 respondents is sufficient for pilot testing in a survey.
Accordingly, the questionnaire items were adjusted for reliability and validity based on the pilot results and the AHoDs suggestions. All the proposals that were advanced by the AHoDs were considered. They included shortening of the questionnaire by omitting redundant and overlapping items, simplifying items, summarizing items that appeared verbose; reconstructing poorly constructed items and elimination of ambiguities. Pilot study, according to Orodho (2005), helps determine both the validity and reliability of the research instruments. The AHoDs used in piloting were ultimately excluded from the main study.

Data Collection Instruments
The questionnaire is the heart of a survey operation (Kothari and Garg, 2014). Semi-structured questionnaires were used to collect data from the AHoDs. The questionnaires were divided into three parts as follows: Part A collected such data on the demographic profiles of respondents and departments. Part B collected data on recruitment practices while part F collected data on lecturers' performance.

Reliability and Validity Tests
Cronbach's alpha coefficient analysis (with the aid of Statistical Package for Social Sciences (SPSS) software) was used to examine the internal consistency of the measures as it is the most reliable test of inter-item consistency reliability for Likert scaled or rating scaled measures (Whitley, 2002& Robinson, 2009. The instrument had adequate reliability for the study with all the items hanging out well with others in all the sub-scales. They were all within the acceptable range of 0.70 to 0.9 as recommended by Tavakol and Dennick (2011) and George and Mallery (2003). Generally, the instrument was of an acceptable reliability standard with an overall scale reliability of 0.796 and a standard deviation of 0.06. Content validity was ascertained by literature search to ensure that the items in the questionnaire were within the domain of the study concepts (Kimberlin & Winterstein, 2008;Drost, 2011) and corroborated by a panel of experts (Cooper and Schindler, 2008;Drost, 2011;Aila & Ombok, 2015). Construct validity, on the other hand, was conducted using Pearson Product Moment Correlations with the aid of SPSS version 21.0. Based on the significant values obtained by the significance (2-tailed)<.05, it was evident that all items in each sub-scale significantly correlated with their sub-scale totals, except for item 4 in the forecasting future requirements and action planning sub-scale. The item was excluded from the final analysis of the data. Generally, the questionnaire items were of adequate validity since they significantly correlated with their total subscales.

Data Analysis
The collected data were processed and organized for statistical analysis. Both descriptive and inferential statistics were used. Pearson's product moment correlation was used to determine the magnitude and direction of relationship between variables (Cooper and Schindler, 2008) while multiple regression used to predict the influence of recruitment practice on lecturers' performance. All tests of significance were computed at α=0.05. The Statistical Package for Social Sciences (SPSS) version 21.0 aided data analysis.

Model Specification
The study sought to establish the influence of recruitment practice on lecturers' performance in Kenyan public universities. This objective was investigated by testing the null hypothesis that: recruitment practice has no influence on lecturers' performance in Kenyan public universities. Recruitment practice was measured using subscales, namely: examination of vacancy; locating, making contact and attracting candidates. The data for each of the two sub-scales was collected using a 5-point Likert scaled questionnaire. The Likert scale responses were then converted in continuous scale data by computing the mean response in each item. The multiple regression model in Equation 1 that follows was used to explore the hypothetical influence of the two aspects of recruitment practice on lecturers' performance in Kenyan public universities. The model was adapted from Cooper and Schindler (2008).

Assumptions of Regression Analysis
Most statistical tests rely upon certain assumptions about the variables used in the analysis (Osborne and Waters, 2002). Consequently, diagnostic tests were conducted on the collected data to confirm their suitability of for multiple regression analysis (Field, 2013). Normality was tested using Shapiro-Wilk's test (Thode, 2002;Gravetter & Wallnau, 2000;Oso and Onen, 2013;Zar, 1999). Durbin Watson test was done to examine independence of errors at a range of 1.50 -2.50 (Tabachnick and Fidell, 2001;Keppel and Zedeck, 1989). The assumption of Multicollinearity was investigated using Tolerance and Variance Inflationary Factor (VIF) (O'Brien and Robert, 2007). Homoscedasticity and heteroscedasticity was examined using scatter plot. Independence of errors was examined using Durbin-Watson (Kothari and Garg, 2014;Fox, 1991). Appropriate action was taken in cases of violations of the assumptions.

Descriptive Statistics on Lecturers' Performance
Lecturers' performance was summarized in four sub-scales, namely: quality of teaching and learning, research and publication, administration and responsibilities, and community engagement and other contributions. Table 2 indicates the summary of lecturer's performance ratings using means and standard deviation.   Kara, Tanui and Kalai (2020) who established that a majority of lecturers in Kenyan public universities had the desired professional quality and engaged in quality instructional practices.

Descriptive Statistics on Recruitment Practice
The two sub-scales of recruitment practice were summarized using means and standard deviations. The findings are presented in Table 3.
European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021  Source: Survey data (2020) From Table 3, it is evident that public universities in Kenya have above average rating in their recruitment practices. This was reflected by an overall mean rating of M=4.02(SD=0.77) in a scale of 1 to 5. This reflects the respondents' satisfaction with recruitment practice in their universities. Specifically, the dimension of Examination of Vacancy received the highest rating (M=4.58, SD=0.53), suggesting that many of the universities do better in this aspect as compared to Locating, Making Contact and Attracting Candidates (M=3.47; SD=1.03).

Relationship between Recruitment Practice and Lecturers' Performance in Kenyan Public Universities
The study sought to establish the relationship between recruitment practice and lecturers' performance. To this extent, Pearson product moment correlation analysis was used to find out the magnitude and direction of the relationships between the individual aspects of recruitment practice and lecturers' performance as well as that of the overall recruitment practice and lecturers' performance, as shown in Table 4. Source: Survey data (2020) It is evident that all the aspects of recruitment practice are directly correlated to lecturers' performance. The findings indicate that the strongest relationship was between examination of vacancy and lecturers' performance, as was reflected by a significant positive correlation coefficient r=.351(p=.000). Comparatively, Locating, Making contract and Attracting Candidates, recorded a relatively weaker positive correlation with lecturers' performance. It was significant at (r=.188, p=.029).
Overall, the correlation between recruitment practice and lecturers' performance was established to be positive and significant (r=.324, p=.000). This suggests a positive association between the two variables, such that when the overall recruitment practice is improved then there would be a corresponding positive improvement in lecturers' performance. It was, therefore, concluded that recruitment practices in Kenyan public universities is positively associated with lecturers' performance. The findings of the study are in conformity with those of Saifaislam, Osman and AlQudah (2014) Table 5.

The Influence of Recruitment Practice on Lecturers' Performance in Kenyan Public Universities
To investigate the influence of recruitment practices on lecturers' performance in Kenyan public universities, the null hypothesis that "Recruitment practice has no influence on lecturers' performance in Kenyan public universities" was tested. This was done using multiple regression analysis, with the investigated null hypothesis being H0: β1=β2=0 and the corresponding alternative hypothesis being H1: at least one βi ≠0. If the null hypothesis is true, then from E(Y)=β0+βi=1,2Xi=1,2 the population mean of Y is β2 for every X value, which indicates that X (recruitment practice) has no influence on Y (lecturers' performance) in Kenyan public universities, and the alternative being that X (recruitment practice) influences Y (lecturers' performance) in Kenyan public universities.
Mean response across a set of statements of Likert-type scale responses in the measure of recruitment practice was European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021 computed to create an approximately continuous variable, within an open interval of 1 to 5 as determined to be suitable for the use of parametric data by Johnson and Creech (1983) and Sullivan and Artino (2013). High scale rating was interpreted to imply a corresponding high implementation of recruitment practice that would influence lecturers' performance. The priori significance level was set at 0.05, such that if the p-value was less than 0.05, then the null hypothesis would be rejected and conclusion reached that recruitment practice has a statistically significant influence on lecturers' performance. If the p-value was greater than 0.05, it would be concluded that a significant difference does not exist. Table 5 shows the results of the regression model. The model summary shows that recruitment practice, as measured by Examination of Vacancy, and Locating, Making Contact and Attracting candidates, accounts for 11.6% (Adjusted R 2 =.116) of the variation in the lecturers' performance in Kenyan public universities. This implies that 11.6% of the variability in performance among lecturers in public universities in Kenya is explained by the combined changes of the two dimensions of recruitment practice: Examination of Vacancy, and Locating, Making Contact and Attracting Candidates. However, the two dimensions had a fairly weak joint correlation (R=.360) with lecturer's performance. On the other hand, the ANOVA output results confirm that the level of recruitment practice is really a significant predictor of performance among lecturers in Kenyan public universities, F (2, 133)=9.890, p=.000. The low value of F suggests that the variability within each of the dimensions of recruitment practice and lecturers performance was generally smaller than variation between the variables. This indicates that the level of recruitment practice in Kenyan public universities can significantly predict lecturers' performance.
It is also evident that the two dimensions of recruitment practice reflected different levels of influence on lecturer's performance. On the one hand, Examination of Vacancy had a significant p-value (t=3.795; p=.000) unstandardized coefficient value of 0.484 within a 95% C.I (.232, .737). On the other hand, the coefficients value for Locating, Making Contact and Attracting Candidates was not significant (B=-0.106; t=.959, p=.339). Nonetheless, given that one of the two dimensions had significant un-standardized co-efficient value, there is sufficient evidence to reject the null hypothesis (H0: β1=β2=0) because at least one βi≠0. Hence, the alternative hypothesis was supported and it was concluded that recruitment practices has statistically significant influence on lecturers' performance in Kenyan public universities.  Vol.13, No.20, 2021 total R squared value for the model (11.6 per cent explained variance) because overlaps or shared variance were removed. Nonetheless, it was concluded that the recruitment practice regression model was adequate to predict lecturers' performance among public universities in Kenya. The model was statistically significant F(2, 133)=9.890, p=.000, accounting for 11.6% (Adjusted R 2 =.116) of the variation in lecturers' performance in Kenyan public universities.
The findings of the study affirm many theoretical arguments, for instance the decision-making theory (Simon, 1945;Mintzberg, 1973;Iyayi, 2002) which is the overarching theory, as well as the findings of other previous studies. Decisions are made in the areas of recruitment with specific reference to strategies to be adopted in recruitment and retention of employees for performance (Gberevie, 2010). Decision-making theory is premised on the fact that decision-making is at the very heart of business success in any organization (Gberevie, 2006). Decision-making theorists hold that decisions are the selection of a proposed course of action (Butler, 1992;Iyayi, 2002). With regards to the influence of recruitment on lecturers' performance, it can be argued that prudent decisions on the construction of valid job descriptions and specifications; identification of the most appropriate sources of candidates; identification of the best media to publicize vacancies; determination of the deliberate strategies that would yield the best candidates among others would lead to the attraction of sufficient and competent pool of applicants from which the best fit can be obtained, thus optimum performance.
Several empirical studies reiterate the influence of recruitment practice on employees' performance. In Afghanistan Civil Service, Rahmany (2018) established that recruitment process, among other variables, accounted for 95% variability in employees' performance. Much as the study ignored recruitment facets such as: examining vacancy, and locating, making contact and attracting candidates (Graham and Bennett, 1998) that the current study addressed. The variation it is greater the 11.6% (Adjusted R 2 =.116) of the variation in lecturers' performance in Kenyan public universities as is in the current study. This could be because of the former examining the influence of recruitment alongside other variables.
In Kenya, via linear regression, Murage, Sang and Ngure (2018) found that recruitment and selection accounted for 14.6% of employees' performance in public universities in Nyeri County. The focus was, however, on ethical issues in recruitment and selection as opposed to the facets of recruitment such as: examining vacancy, and locating, making contact and attracting candidates which characterize the current study (Graham and Bennett, 1998). Moreover, the unit of analysis was non-teaching staff and not academic staff as in the current study. This would probably create disparities owing to different measures of performance used to gauge the two categories of university employees. Furthermore, the contribution of recruitment and selection were not established in isolation.
Examining the influence of recruitment and selection on the performance of employees in research institutions in Kenya, Ombui, Mukulu and Waititu (2014) established that recruitment and explained 37.4 of variation in employees' performance. The unique contributions of recruitment and selection were not established. Furthermore, the descriptors of recruitment were not revealed. The other studies that demonstrated a positive influence of recruitment on employee performance are Mokaya, Mukhweso and Njuguna (2013) and Suntanto and Kurniawan (2016). According to the latter, Recruitment accounted for 53.11% variation in performance in Batik Industry in Solo City, Indonesia. However, just like the other studies, they ignored the facets of recruitment such as: examining vacancy, and locating, making contact and attracting candidates which characterize the current study (Graham and Bennett, 1998).
On the flip side, examining the effect of recruitment, selection and placement on employee performance at PT Green Glovers in Indonesia,  established that recruitment had no effect on employee performance. However, the study relied on a saturated sample of only 90 respondents in a one company which may not be suitable for generalization.

Conclusion
This study was aimed at establishing the influence of recruitment practice on lecturers' performance in Kenyan public universities. In conformity with the results of many previous studies, correlation results have revealed a statistically significant positive correlation with lecturers' performance. Similarly, regression analysis has revealed that that enhancement of recruitment practice can improve lecturers' performance. Consequently, it is recommended that public universities invest in recruitment practice, owing to its benefits. European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.20, 2021