METHODOLOGY ON EFFECTS OF HUMAN RESOURCE PRACTICES

METHODOLOGY ON EFFECTS OF HUMAN RESOURCE PRACTICES ON

EMPLOYEE TURNOVER IN THE FLOWER INDUSTRY

Research Design

Creswell (2009) defines research designs as plans and the procedures for research that span the

decisions from broad assumptions to detailed methods of data collection and analysis. Research

design represents the arrangement of conditions for collection and analysis of data with the aim

of combining relevance to the research purpose with economy to the procedure so that the

questions regarding decisions on what, where, when, how much and by what means concerning

the research study constitute a research design (Orodho, 2008). This study adopted survey design

both analytical and descriptive. According to Sapsford (2006) a survey is a detailed and

quantified description of a population. Surveys involve the systematic collection of data, whether

this is by interviews, questionnaire, or observation methods so at the very heart of surveys lays

the importance of standardization.

This research was to investigate factors influence employee turnover in flower farms in North

Rift, Kenya. Hence, a survey research design was considered appropriate for the study

because the data was collected from cross sectional representation of the population. The views

obtained therefore represented the views of the entire population under study. Information was

collected from respondents about their experiences and opinions in order to generalize the

findings to the population that the sample was intended to represent. This method was the most

appropriate for obtaining factual and attitudinal information for research questions about

opinions, characteristics and present or past behavior (Neuman, 2000). Since this study sought to

obtain information from flower farms employees about human resource practices the survey

design was the most appropriate.

Research Paradigm

According to Bryman (2004), a paradigm is a set of beliefs that prescribes how research in a

specific discipline should be implemented and how the results should be interpreted. According

to Tashakkori & Teddlie, 2010 a paradigm is essentially a set of beliefs that encompasses the

theories of a group of researchers, with these ideas underpinning their research methods and

interpretation. Within the context of social and behavioural sciences, paradigms largely follow

two main approaches. Guba and Licoln apply the terms scientific and naturalistic to the concept

of a paradigm, while Tashakkori and Teddlie (2010) use the terms positivist and constructivist.

The nature and underlying understanding of these two theories of the paradigm have been a great

source of debate. For example, Burrell and Morgan (1979) discuss the influence that a paradigm

position can have on the ensuing design approach; a largely quantitative research method can be

indicative of underlying positivist paradigm beliefs, while a qualitative approach is more

suggestive of a constructivist paradigm position. According to Bryman, while these types of

research design may be indicative of the nature of the paradigm position, they are not fixed

(2004).

In their studies, Tashakkori and Teddlie (2010) address the way in which debates relating to

paradigms have evolved throughout history, and assert that the once‐dominant logical positivism

that was founded on the need for observable facts became less popular with the emergence of the

popularity of the post‐positivist position. Essentially derived from the underlying beliefs of

positivism, post‐positivism acknowledged the theory‐ladenness of observation, ladenness of

facts, and the value‐ladenness of science and research as being constructivist in nature

(Tashakkori & Teddlie, 2010).

As these theories developed they gradually were superseded by constructivist perceptions of

social reality (Tashakkori & Teddlie, 2010) during what was known as the mono method era.

During this period researchers began to confine their studies to either a quantitative or qualitative

approach, underpinned by their post‐positive or constructivist beliefs respectively. Paradigm

Wars according to (Tashakkori & Teddlie, 2010) describe how the mono‐method approach to

scientific study was challenged in the 1960s, with the resulting approach being a mixture of

qualitative and quantitative research and thus a hybrid of post‐positivism and constructivist

beliefs.

In the 1990s, mixed‐method approaches that encompassed both qualitative and quantitative

studies became highly popular (Creswell, 2003). The period of paradigm wars marked the

emergence of debate that focused on the ways in which paradigm and methodology are related

(Tashakkori & Teddlie, 2010). Some scientists viewed the differences between post‐positivism

and constructivism as being entirely conflicting, and did not agree to the viability of mixed

method research. They became known as the compatibility theorists in response to this, a third

set of beliefs eventually emerged: the pragmatic paradigm. Many pragmatists, like Tashakkori

and Teddlie (1998), believe that the pragmatic paradigm approach has resolved the

disagreements of the paradigm wars; and subsequent studies have employed a mixture of

qualitative and quantitative research to good effect (Meekers, 1994; Morse, 1991).

Pragmatists’ associate paradigm with the nature of the research questions developed (Creswell,

2003). As it is not possible to conduct research in a single‐dimensional approach, a ‘what works’

tactic allows researchers to focus on questions that cannot be handled via a purely quantitative or

purely qualitative approach. Darlington and Scott (2002) add to this argument, asserting that

many decisions that are made while designing a research study are not based on decisions that

are philosophical in nature, but on practical concerns related to the type of methodology that will

be best suited to the study. Essentially, therefore, the pragmatic paradigm rejects a strict choice

between post positivism and constructivism (Creswell, 2003).

Tashakkori and Teddlie (1998)) and Creswell (2003) view the pragmatic paradigm as a

philosophy that supports an intuitive approach to research and scientific study. Adopting this

approach allows scientists to make informed decisions about which methods to apply, based on

their individual value systems (Creswell, 2003). As such, it may be claimed that the pragmatic

paradigm is suitable for social and management research as well as for scientific research, as it

offers a harmonious combination of quantitative and qualitative approaches suitable for

practitioner‐based research. This research was guided by the philosophy of pragmatism.

Pragmatic research reflects the researcher’s innate disposition toward systematic enquiry. It also

allows a flexible approach to the investigation, accommodating an outcome‐and adaptive‐

oriented enquiry method (Johnson & Onwuegbuzie, 2004) and allowing for the use of

mixed methods (Calori, 2000; Cherryholmes, 1992; Creswell, 2003). This kind of

developmental and iterative approach helps the researcher to engage with issues as they emerge.

In addition, it allows for both qualitative and quantitative data and analysis to be used, and both

inductive and deductive forms of enquiry.

Research Approach

A research project that employs both qualitative and quantitative techniques can be said to be

using a mixed method approach. This approach incorporates different types of data to help in

better answering the research questions (Hayati, Karami, & Slee, 2006). It has been suggested

that a mixed method approach is best suited to exploratory research, as the questions being posed

have not been answered before (Gable, 1994; Karami, Analoui, & Rowley, 2006; Scandura &

Williams, 2000). A mixed method approach adds to the credibility of outcomes as the

quantitative data is supported by qualitative data (Easterby‐ Smith et al.,Easterby‐Smith,

Thorpe., & Lowe, 1991; McGrath, 1982; Scandura & Williams, Employing both qualitative and

quantitative techniques brings a further perspective to the research questions. As Punch (2005)

contends, qualitative techniques help in determining the attitudes, behaviours, and perspectives

of the research subjects while quantitative techniques help in understanding the environment of

the study. When combined, the methods will present a lucid picture and will offer clear answers

to the research questions. Greene, Caracelli and Graham (1989) discuss five benefits of

employing a mixed‐methods approach: Triangulation using different sets of data and

methodology in order to test hypotheses and consistency of findings; Complementary

confirming the validity of the results from one study by employing a different research method;

Development applying the results from one method in the design of further research; Initiation

challenging research results from one method; Expansion developing methods in order to explore

them further and garner additional detail. It is generally recognized that a mixture of qualitative

and quantitative methods provides the most reliable insights and research findings.

Study Area

The study was undertaken in flower farms in North Rift Region of Kenya. North Rift Kenya is

positioned in the Northern part of the former Rift Valley Province. It comprises of several

counties, namely: Nandi, Uasin-Gishu, Trans Nzoia, Elgeiyo Marakwet, West Pokot, and

Turkana. The main economic activity is agriculture which includes floriculture. This area is

predominantly affected by perennial conflict and insecurity due to scarce natural resources,

livestock rustling, cross border banditry, land disputes, ethnic rivalries and proliferation of illegal

small firearms. These issues are further compounded by the poor communication infrastructure

in the region. During the time the study was conducted it was relatively peaceful thus enabling

the researcher to complete the research in time. The area covers 91375.4 square kilometres and

has a population of 4,203,988. This study however focussed on six flower farms four were in

Uasin Gishu and two were in Trans Zoia. The researcher selected these farms for study due to the

following; accessibility, budgetary constrains, the farms employed both men and women so

information obtained will represent the whole population rather than one predominant gender.

Flower farms in North Rift, Kenya are relatively a new development in the area hence the

researcher had an interest to establish how well human resource practices were being

implemented. This is therefore the first research of its kind in the flower industry dealing with

human resource practices.

Target Population

Farm Managers Supervisor Employees Total

Majimazuri Equator

flowers Zena Roses

Panacol Flowers

Anderson flowers

Sirgoek

970

772

975

825

826

476

990

788

989

835

836

484

Total 30 48 4844 4922

Population is a large collection of all subjects from where a sample is drawn (Zikmund, Babin

Carr & Griffin, 2012). The target population or the unit of observation is a group of individuals,

or objects that a sample is drawn for measurement (Kombo & Troomp, 2009). The study targeted

employees and management in six flower farms in North Rift Kenya. The farms have a total of

4922 employees as shown in Table 3.1.

Sample size and Sampling Procedures

Sample size determination

Krejcie and Morgan (1970) as cited by Kasomo (2001) based the sample size for this study on a

sample size determination formula. The sample size calculation is in Appendix IV. The total

sample size for the study was as shown in Table.

Sample Size

Sam

plin

Procedures

Sampling procedures are the plans or strategies used by a researcher to select a sample of participants

Farm Managers Sample

Size

Supervisor Sample

size

Employees Sample

size

Total

Sample

Majimazuri

8 1

970 70 72

Zena Roses

6 1

772 55 57

Equator flowers

7 1 7 1

975 69 71

Panacol Flowers

3 1 7 1

825 58 60

Anderson flowers

4 1 6 1

826 59 61

Sirgoek

2 1 6 1

476 34 36

Total Target 30 48 4844

Total Sample size 6 6 345 357

chosen from a given population to gain or obtain information about the large group, (Cresswell, 2011).

Employees were stratified into 3 groups (top management, supervisor and employees. The study used

purposive sampling to select participants from the top management employees.. Cresswell, (2011) state

that purposive sampling which is also known as deliberate or convenience sampling allows researchers to

use cases that have the required information with respect to the objectives of their study.

Stratified sampling was used to select employees from the middle level management, supervisors and

lower cadre employees. Simple random sampling was applied to select the actual participants for the

study from each flower farm. Under Simple random sampling, each member of a population has an equal

chance of being included in the sample. Also, each combination of members of the population has an

equal chance of composing the sample. Simple random sampling is the easiest method of sampling

and it is the most commonly used. Advantages include additional information on the frame such as

geographic areas other than the complete list of members of the population along with information for

contact.

Also, since simple random sampling is a simple method and the theory behind it is well established,

standard formulas exist to determine the sample size, the estimates and these formulas are to use

(Stebbins, 2001). On the other hand, this technique makes use of auxiliary information present on the

frame that could make the design of the sample more efficient. And although it is easy to apply simple

random sampling to small populations, it can be expensive and unfeasible for large populations because

all elements must be identified and labeled prior to sampling. It can also be expensive if personal

interviewers are required since the sample may be geographically spread out across the population,

(Silverman, 2011). However, the method was still appropriate for the study since the study sample was

not so large. This implied the selection by the researcher, of participants for a particular study he/she

deems in the best position to provide the relevant information needed for such a study. This section

provides the sampling process adopted for this study. The study adopted stratified sampling to select

employees where respondents were grouped in different farms then random sampling employees in every

stratum. Stratified sampling identified sub groups in the population and their proportion and selected from

each group to form a sample. This was intended to group the flower farms and their population into

homogenous subsets that share similar characteristics. It also ensured equitable representation of the

population in the sample (Oso and Onen, 2009). Purposive sampling was also adapted to select managers

and supervisors who participated in the study in every farm.

Data Collection Instruments

Data collection instruments are means by which primary data is collected in social research (Kothari

2009). The methods are varied in terms of time, cost of money or other resources at the disposal of

researcher (Orodho, 2008). The study used questionnaires and observation to collect primary data.

Questionnaires consist of specific short questions related directly to the research questions (Copper and

Schindler, 2011) which are asked verbally by the interviewer or answered by the respondents on their

own and the number of the closed ended questions should always exceed the open ended (Bryman,

2012). This study used structured questionnaires which were self administered to obtain the primary data.

The items were adopted from other relevant studies to ensure consistency and flow of survey was

preserved. Secondary data was collected trough document analysis.

Data Collection Techniques

This study utilized quantitative and qualitative data. According to Creswell (2009) research involves

collecting and analyzing quantitative and qualitative data. Quantitative data includes closed-ended

information such as that found on attitude, behavior, or performance instruments. Sometimes quantitative

information is found in documents such as census records or attendance records. The analysis consists of

statistically analyzing scores collected on instruments, or public documents to answer research

questions or to test hypotheses. This section presents the data collection instruments that the study

adopted. These Instruments are; questionnaires, observation and document analysis, as discussed below.

Primary Data

According to Donald Currie (2005) Primary data is data that are only obtainable directly from an original

source. In certain types of primary research, the researcher has direct contact with the original source

of the data. The decision to collect primary data for a research project is influenced by the kind of

research that the searchers are carrying out. A researcher carries out primary research when the data you

need is not available from published sources. For example, if you are carrying out an assignment, a major

project or a degree dissertation, you may need information that is only available from key individuals,

such as managers, a group of employees in an organisation, customers or other members of the public.

Conversely, you may need to know how groups and individuals react to particular situations and ideas, or

how they behave when they are carrying out their jobs. There are two main methods used to collect

primary data in this research. The methods were questionnaires and observations.

Questionnaires

A questionnaire is a set of questions on a topic or group of topics designed to be answered by a

respondent. This implies that the respondent is in full control of the questionnaire and will thus complete

and return it at their on convenient time. According to Kombo& Tromp (2006), a questionnaire is a

research instrument that gathers data over a large sample. The questionnaire was suitable for this study,

mainly because the variables under study could not be directly observed such as views, opinions,

perceptions feelings and attitudes of the respondents. Such information is best collected through

interviews, (Kothari, 2008). Since it was a standard research instrument, it allowed for uniformity in the

manner in which questions were asked and made it possible to compare across respondents, (Cohen and

Manion, 2003). It was also suitable because the target population was literate and thus there were limited

chances of difficulties in responding to questionnaire items.

However, a questionnaire has some disadvantages. For example it may not gauge the respondents’

sincerity; it might not reach the targeted destination, and is not useful when the target population is

illiterate. The questionnaire contained closed to elicit attitudes, views, perceptions and opinions from the

subordinate staff. The items were in 5 point Likert scale to enable the researcher to code responses easily

using descriptive statistics. Questionnaires were administered by the researcher. They were constructed

by researcher and approved by supervisors. According to Creswell (2009) the questionnaire is an efficient

research tool because the researcher is likely to obtain personal ideas from the respondent. The

respondents are free to read questions, complete the questionnaire and return it at the end of the activity.

The questionnaire formed the major source of primary data used in the study.

Gillham (2007) points out, the popularity of questionnaire is also probably based on some of their

inherent advantages. They are low cost in terms of both time and money. In contrast to, say, interviews,

questionnaires can be sent to hundreds or even thousands of respondents at relatively little cost. The

inflow of data is quick and from many people. Also respondents can complete the questioners at a time

and place that suits them. Data analysis of closed questions is relatively simple and questions can be

coded quickly. Respondents’ anonymity can be assured, but Gillham (2007), rightly noted that in small

scale surveys, this can be largely nominal in character. It may not be difficult for the researcher to

recognize the responses of individuals. But real anonymity can also be double edged. The data from

questionnaires was purely quantitative and all the objectives were covered; section A sort information on

demographic profiles on the respondents. Section B delt with management styles, section C delt with

intrinsic and extrinsic rewards, section D work life balance and finally section E health and safety.

Observation

Observation involves the systematic viewing of people in action and the recording, analysis and

interpretation of their behaviour (Saunders et. al., 2012). The observational method is often associated

with ethnographic methodology in that it studies people in their natural settings. There was need to

employ naturalistic observation methods in gathering the data so that the researcher may observe the

employees in their natural phenomenon. An essential ingredient of naturalistic observation is careful

record keeping (Burney, 1998), particularly the use of observational schedules. Observation method was

used because the study took place in the flower farms where employees were in their natural setting. This

was an important method that would help to find the comparison between the data collected from the

questionnaire and the reality in the flower farms. The data obtained from this section covered all the

objectives of the study and the results were integrated with the quantitative results from the questionnaire

in chapter four.

Secondary Data

Document Analysis according to Lindlof and Taylor (2002) can be linked to the talk and social action

context that the researcher is studying. Documents are not only stable, exact and broad, but are also

reliable because they are not created as a result of, or to suit, the research context (Yin 2009). He argues

that documents help researchers reconstruct past events as well as ongoing processes that are often

relatively accurate and documents reflect a certain kind of rationality at work. Blaxter et. al., (2001)

says that using documents may confirm, modify or contradict a researchers finding, enable a researcher to

focus attention on analysis and interpretation and also complement primary data. Lindlof and Taylor

(2002) started that common documentary sources for research are government reports; flower farms

records, employee unions, Federation of Kenyan Employers.

This was an important method of data collection because the study reviewed the various literatures that

have been written by various authors. This will become the background of the available information

where the study will establish new findings that foster new knowledge to flower farming. However,

secondary data has advantages and setbacks. Its advantages are; Cost effective. This is likely to be less

than would have been incurred if the data is collected from scratch. The researcher can get data without

necessarily going to the field to obtain primary data. If the data is publicly available, then scholars will

have the opportunity to carry out replication studies to fine tune or validate initial findings (Welch,

2000).Data often come from sources developed by teams of researchers who have many years of

experience in research design and data collection (Boslaugh, 2007).Convenience for student research this

was suggested by (Szabo & Strang, 1997).

Given that they often have to complete dissertations to very demanding timelines. It also has some

disadvantages. Data might be incomplete, obsolete, inaccurate, or biased. In case of this set back a

researcher can use unpublished studies. Data mining researcher may go round that data by looking for

themes of hypotheses. The data could be split into two separate sub-samples, the first used for

exploration and the second for hypothesis. The data obtained from this section focused on all the

objectives and the results were integrated with the quantitative results from the questionnaire in chapter

four.

Validity and Reliability of the Research Instruments

This section deals with validity and reliability of the research instruments.

Validity of the Study Instrument

Validity is the degree to which an instrument measures what it is expected to measure (Bryman, 2012). It

is also regarded as the extent to which the differences found with a measuring instrument reflect true

differences among those tested (Kothari, 2011). There are various types of validity that include

construct, content, convergent concurrent, and predictive (Drost, 2011). This study adopted both the

construct and content validity. The construct validity was specifically applied by use of a questionnaire

that was divided into sections to collect information for each of the objectives and also taking into

considerations the conceptual framework (Nasrudin & Othman, 2012) while the content validity was by

KMO and Bartlett's Test

Objective One

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .712

Approx. Chi-Square 136.197

Bartlett's Test of Sphericity Df 55

Sig. .000

Objective Two

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .506

Approx. Chi-Square 656.283

Bartlett's Test of Sphericity Df 351

Sig. .000

Objective Three

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .502

Approx. Chi-Square 592.893

Bartlett's Test of Sphericity Df 276

Sig. .000

Objective Four

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .909

Approx. Chi-Square 677.717

Bartlett's Test of Sphericity df 30

Sig. .000

Dependent Variable :

employee turnover: Indicators: dismissal, Attrition and resignation

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .772

Approx. Chi-Square 63.083

Bartlett's Test of Sphericity Df 15

Sig. .000

achieved by adopting and revising instruments that had been used in earlier studies and subjecting the

questionnaire to experts who are a panel of peers to assess whether the questions are effective

(Bryman, 2012). According to Foxcroft (2004), a panel of experts to review the test specifications and the

selection of items, the content validity of a test can be improved. The experts were able to review the

items and comment on whether the items cover a representative sample of the behaviour domain. To test

the validity of the instruments used in the study, the questionnaire was availed to supervisors together

with a panel of experienced researchers of Moi University. The results from the piloting together with the

comments from the experts were incorporated in the final instrument. Also factor analysis was employed

to check on the factors to be reduced.

Validity factor analysis

Kaiser Meyer-Olkin and Bartlett Test of Sphericity

Table Kaiser-Meyer-Olkin and Bartlett’s Test Results

To assess the factorability of items, two indicators were examined Kaiser Meyer-Olkin Measure of

Sampling Adequacy and Barletts Test of Sphericity. These tests were generated by SPSS, and helped to

assess the factorability of data or suitability of data for structure detection (Pallant, 2010). Kaiser-Meyer-

Olkin (KMO) test was used to assess sampling adequacy. The index ranges from 0 to 1 (Tabachnick &

Fidell, 2007). For adequate sample, KMO test statistic should be greater than 0.5 (Hair et al., 2013). The

world-over accepted index is 0.6 or higher to proceed with factor analysis (Fabrigar, Wegener,

MacCallum and Strahan, 1999). Table 3.3 shows KMO statistics of 0.712, 0.506, 0.502, 0.909 &0.772

for objectives one, two, three, four and dependent variables which is greater than the conventional

probability value of 0.5 and over .60 for a satisfying sample. This implies an acceptable degree of sample

adequacy for factor analysis.

On the

other

hand Table 3.3 also presents the results of Bartlett’s test of sphericity. Bartlett test of sphericity was

performed to assess the appropriateness of using factor analysis (Hair, et al., 2013). For factor analysis to

be recommended suitable, the Bartlett’s test of sphericity should have p-value of less than 0.05 (Fabrigar

et al., 1999). Bartlett’s test of sphericity indicates a chi-square of 136.197, 656.283, 592.893, 677.717 &

63.083 with an associated p-value of 0.00 which is lower than the conventional probability value of

0.05. It was thus concluded that factor analysis was an appropriate approach for assessing construct

validity of the scale.

Reliability of the Instrument Table Reliability of the Instrument

Reliability Statistics

Reliability is the consistency of measurement (Bollen, 1989; Abbot and McKinney, 2013) despite the

changing conditions. There are a variety of methods that can be used to test reliability in behavioral

research that include; test-retest reliability, alternative forms and split halves and internal consistency

(Drost , 2011). The internal consistency was selected due to its higher stability in comparison to the others

(Bryman, 2012). And tested using the Cronbach’s alpha statistic (1951) which measures consistency

within the instrument assessing how well a set of items measures a particular behavior or characteristic

within the test. The estimates of reliability should be based on the average inter correlations among all the

single items in the test (Drost, 2011). The Cronbach’s alpha (α) coefficient when used for reliability test

the value should be 0.8 although 0.7 can be adopted as a satisfactory level (Bryman, 2012). In this study a

Cronbach alpha of 0.829 was obtained indicating that the instruments were reliable Pilot test was

administered on 30 respondents at Ravine Roses in Baringo County. After which Cronbach’s alpha was

used to present the average of all possible split-half correlation and so measures the consistency of all

Cronbach's Alpha N of Items

.829 101

items and the results were 0.829. The extent of this consistency is measured by reliability consistency

using a scale from 0.00 which is(very unreliable) to 1.00 (perfectly reliable). A score of greater than 0.7 is

generally deemed to be acceptable (Gray 2014). The results are presented on the table above.

Measurement and Scaling Technique

Measurement is the assignment of a number to an object which reflects the degree of possession of a

characteristic by that object (Panneer selvam, 2006). Scaling is a description of procedures of assigning

numbers to degree of opinions, attitudes and other concepts (Kothari, 2011). The study used closed

ended questions and a 5-point Likert scale for measuring the objectives. The closed ended questions

presented on the Likert scale is designed to examine how strongly the subjects agree or disagree with a

statement (Sekaran and Bougie, 2010). The 5-point Likert will range from strongly disagree to strongly

agree or from highly significant to highly insignificant. Likert scaling is one- dimensional (Trochim,

2006) and is usually preferred due to the fact that the concepts are easy to understand as one has more or

less of it; they are reliable and provide more information (Kothari, 2009). The measurement and scaling

technique of using the Likert was applied in the study.

Measurement of Independent and Dependent Variables

The Likert scale dominated the questionnaire due to the fact that it can be used in a wide variety of

circumstances: when the value sought cannot be asked or answered with certainty ;when the value sought

is an opinion, effect or belief or when the value sought is considered so sensitive in nature that the

respondents cannot answer except if it is in large ranges and the fact that it can easily be evaluated

through standard techniques such as stepwise regression analysis and factor analysis (Montgomery, Peak

and Vining, 2001).

Data Analysis and Processing

Table Data Analysis by Objective Measurement and Test

OBJECTIVE Nature of

objective

Independent

variable

Dependent

Variable

STATISTICAL

TESTS

Management commitment

and employee turnover

Inferential

descriptive

Non metric Metric

Correlation Ordinal

Regression,

Rewards and employee

turnover

Inferential

descriptive Non metric Metric

Correlation Ordinal

Regression,

Work life balance and

employee turnover

Inferential

Descriptive Non metric Metric

Correlation Ordinal

Regression,

Health and safety and

employee turnover

Inferential

descriptive Non metric Metric

Correlation Ordinal

Regression,

From Table above data analysis is the application of reasoning to understand data that has been gathered

with the aim of determining consistent patterns and summarizing the relevant details are revealed in the

study (Zikmund, Babin, Carr and Griffin, 2012). Data processing entails editing, classification and

tabulation of data collected so that easy to analyze (Kothari’, 2009). Data entry converts raw data

gathered by secondary or primary methods to a medium for viewing and manipulation. The analysis used

descriptive and inferential statistics. The hypothesized relationship was examined and tested using

Pearson correlation analysis and Regression analysis. Data after collection is processed and analyzed for

the intended purpose (Cresswell, 2011). The processing involved editing and scrutinizing the completed

questionnaire so that data was accurate, consistent with the facts gathered and uniformly entered at the

field and after all the questionnaires had been gathered so that it was arranged for coding and tabulation.

The data was coded through assignment of numbers to facilitate quantitative analysis.

Ethical Considerations

According to Jwan and Ong’ondo (2011) research ethics are the moral principles that guide a

research study from its inception through to its completion and publication. The respondents were assured

of anonymity by asking them not to write their names on the questionnaires. The researcher also took

individual responsibility for the conduct and consequences of the research by adhering to the time

schedule agreed upon with the management of flower farms. The appropriateness and acceptability of our

behaviour as researchers was affected broader social norms of behaviour. Ethical issues in this study were

categorized into three.

Ethical issues relating to Research Process

The researcher maintained privacy of actual and possible participants. Maintenance of the confidentiality

of data provided by individuals or identifiable participants and their anonymity. The researcher respected

the willingness of the despondence to participate not to participate in the research. Ethical issues Relating

to the Individual Researchers. The researcher took step to avoid plagiarism, fraud and abuse of

privileges. Plagiarism was avoided by acknowledging the author. It is also the stealing of ideas from

another scholar Creswell et al (2003). It’s a criminal offence, punishable by law. Fraud is a

situation where the researcher takes data that has been collected. An instance is when a researcher

completes a questionnaire on behalf and in the absence of the respondent Creswell et al (2003). Fraud

was avoided by collecting primary data. Absence or misuse of researcher’s privileges refers to a

situation where the researcher misuses the trust of the subject. The researcher maintained confidence of

the despondence by remaining honest.

Issues Relating To Research Subjects

During the study the researcher upheld the right of the despondence: the right not to participate, not to be

harassed or, not to be contacted at unreasonable time and place, of participants to determine with reason

when they will participate in the data collection process, , of participants not to answer any question, or a

set of questions. To achieve this researcher contacted the participants during lunch brakes and issued

questionnaires to those who were willing to take.

REFERENCES

Abbas, A. (2010). Islamic challenges to HR in modern organizations. Personnel Review, 39(6),

692-711.

Abbasi, S., & Hollman, K. (2000). Turnover: The real bottom line. Public Personnel

Management, 29(3), 333.

Abdalla, I., & Al-Homoud, M. (1995). A survey of management training and development

practices in the State of Kuwait. The Journal of Management Development, 14(3), 14-25.

Abdullah, A& Ramay, M.I. (2012). Antecedents of organizational commitment of Banking sector

employees in Pakistan. Serbain Journal of Management. 7(1), 89-102.

Abdullah, M.M.B., Uli, J. & Tari, J. J. (2008). The Influence of Soft Factors on Quality

Improvement and Performance, The TQM Journal, 20 (5), 436-452.

Abdullah, R., Zain, R., Mushaireen M., Khalid. K & Mohd T.H. (2012). The Effects of

Teamwork towards Jobs satisfaction in Hotel Industry in Klang Valley, Malaysia. International

Journal of Business and Behavioral Sciences, 2, (3) 8-19.

Abeysekera, R. (2007). The Impact of Human Resource Management Practices on Marketing

Executive Turnover of Leasing Companies in Sri Lanka. Contemporary Management Research.

(3) 3. 233-252.

Achoui, M., & Mansour, M. (2007). Employee turnover and retention strategies: Evidence from

Saudi companies. International Review of Business Research Papers, 3(3), 1- 16.

Adams, S. J. (1965. Inequity in social exchange. In: Derkowitz, L. (Ed.). Advances in

Experimental Social Psychology, 2, 267-299. Academic Press: New York.

Agwu, M.O. & Olele, H.E. (2014). Perception Survey of Employees Participation in Decision

Making and Organizational Productivity in Julius Berger Nigeria PLC Bonny Island. British

Journal of Economics, Management & Trade, 4, 620-637.

Agwu, M.O. (2013). Impact of Fair Reward System on Employees Job Performance in

Nigerian Agip Oil Company Limited Port-Harcourt, British Journal of Education, Society &

Behavioral Science.

Akintoye IR (2000). The place of financial management in personnel psychology: a Paper

Presented as Part of Personnel Psychology Guest Lecture Series, Department of Guidance and

Counselling. University of Ibadan, Nigeria.

Akroyd, D., Jackowski, M.B., & Legg, J. (2007). Factors Affecting Radiographers’

Organisational Commitment Radiologic Technology, 78, 467-475.

Alfes, K., Shantz, A.D., Truss, C., and Soane, E.C. (2013). The link between perceived

human resource management practices, engagement and employee behaviour: a moderated

mediation model. The International Journal of Human Resource Management, 24, 330-351.

Alimario, A. (2011). 80% of professionals unhappy with jobs: study. Business Mirror.

Retrieved from cbnnews.com/business/04/28/11/80- professionals-unhappy-jobs- study.

Allan, C., Loudoun, R. & Peetz, D. (2007). Influences on work/non-work conflict’.

Journal of Sociology, 43 (3): 219-239.

Allen R.S., & Helms M.M., (2001). Reward Practices and Organizational Performance”

Compensation &Benefits Review.

Anderson, D. & Kelliher, C. (2009). Flexible Working and Engagement: The Importance of

Choice. Strategic HR Review, 8, (No. 2), 13-18.

Anderson, T.W.(2003). .An Introduction to Multivariate statistical analysis.3rd ed. New Jersey:

John Willey Sons.

Argyle, M (1989) The Social Psychology of Work, Penguin, Harmondsworth

Armstrong, M. (2010). A handbook of Human Resource Management practice, (11th ed.).

London: Kogan page.

Armstrong, M., (2006). Human Resource Management Practice. 10th Edition ed. London:

British Library.

Armstrong, M., (2010). Armstrong's Essential Human

Students also viewed