The motivational value of rewards amongst Malawi ’ s health professionals

Main findings: The study confirmed the existence of reward-related problems amongst health professionals in Malawi’s MoH. Initially, statistical testing of the hypothesised model collapsed, suggesting that no relationship existed amongst the variables. However, statistical testing of the re-specified model suggests that there is a positive relationship between financial and non-financial rewards. Through the structural equation modeling (SEM) exercise, an inverse (negative) relationship between financial and non-financial rewards was established.

Whilst the researcher acknowledges the availability of qualitative research reports revealing the existence of these reward-related problems, no comprehensive, quantitative and scientifically informed study has been conducted to substantiate them.Muula, 2006;Muula & Maseko, 2005;World Bank, 2004) claim that shortage of health professionals in Malawi's Ministry of Health (MoH) is very severe even by African standards.This shortage mainly arises from reward-related problems and threatens the realisation of plans for scaling up interventions to control the spread of diseases in Malawi.

Aim and objectives of the study
The aim of the study was to assess rewards and their motivational value amongst health professionals in Malawi's MoH.The specific objectives of the study were as follows: • To review the existing literature on rewards and motivation for workers with special attention to health professionals in Malawi's public health sector.• To assess the knowledge of a sample of health professionals regarding the rewards system(s) operational in Malawi's MoH.
• To investigate the perceptions of a sample of health professionals about the efficiency and motivational value of the reward system(s) operational in Malawi's MoH.

Value of the study
This research contributes to the theoretical knowledge of reward practices in the public health sector by expanding the body of knowledge about the value of financial and nonfinancial rewards as a means to attract and retain health professionals in the public health service domain.Therefore, various public health sector employers in Africa can use this study to guide the implementation of practices that motivate professionals to commit themselves to public health care services.The findings of this study could influence the decision-making processes regarding the introduction of efficient reward systems in the health sector in particular, as well as other public sector organisations in general.
There are three major schools of thought regarding rewards and motivation which have emerged in the field of industrial and organisational psychology.The first school views financial rewards as a primary motivator (Bowey, 1997;Feber & Nelson, 2009;Gellerman, 1963;Robbins, 2010;Taylor, 1947;Weber, 1947).Financial rewards are a key in recognising workers' achievements and aspirations; reinforce desirable work behaviour; are a powerful means of informing workers what is actually valued or not valued in the organisation; are regarded by workers as highly tangible strategies for recognising their worth; and improve the workers' selfesteem and gain the esteem of others (Robbins, 2010).If financial rewards are inadequate, they can be a major source dissatisfaction and migration amongst workers (Armstrong, 2010;Armstrong & Murlins, 2007;Duncan, Gruenburg & Wallis, 1980;Greene, Caracelli & Graham, 1989;Greenfield, 1993;Holbeche, 2010;Podger, 2002;Robbins, 2010;Stanley & Berger, 1965).Robbins (2010) argues that the motivational value of financial rewards is greater than that of non-financial rewards because of their potential to meet various human needs advocated by Maslow (1943).'People may not work for money [financial rewards], but take away the money [financial rewards], how many people would come to work?' (Robbins, 2010, p.180).In other words, financial rewards are a primary motivator in any employment relationship.Therefore, organisations have an obligation to use financial rewards to attract, motivate and retain workers.Workers use financial rewards to compare job offers since it is easier to compare financial (extrinsic) rewards as opposed to non-financial (intrinsic) rewards.Financial rewards play an important role in workers' decisions to join or leave an organisation (Armstrong & Murlins, 2007;Baron-Punda, 2002;Caudron, 1993;Gellerman, 1963;Robbins, 2010;Taylor, 1947;Weber, 1947;Podger, 2002;Redmond, 2010).This school is backed up by classical theory which asserts that business organisations that have accepted this theory believe that workers are driven by the desire to earn maximum financial rewards.The theory's assumption is that workers are motivated primarily by financial rewards and will maximise their work output if they are offered extra financial rewards for any additional work (Feber & Nelson, 2009;Gellerman, 1963;Robbins, 2010;Taylor, 1947;Weber, 1947).
The second school of thought supports the argument that the use of non-financial rewards as a primary motivator is important because they are effective when assignments require conceptual and creative thinking.Additionally, they are powerful in motivating positive work behaviour and building feelings of confidence and satisfaction amongst workers.Furthermore, the motivational value of non-financial rewards is long term and less costly than that of financial rewards (Cole, 2004;Landy & Conte, 2010;Robbins, Odendaal & Roodt, 2005;Snelgar, 2012).This school of thought is backed up by Herzberg's (1960Herzberg's ( , 1966) ) two-factor theory which describes factors such as achievement, recognition, increased responsibility, growth and development or advancement as motivators.
Factors such as salary and other forms of financial rewards, company policies and administration, supervision, security, interpersonal relations and working conditions have been described as hygiene factors that do not motivate workers.
The third school of thought favours a well-balanced mixture of financial (extrinsic) and non-financial (intrinsic) rewards (depending on the situation and individual reward preferences) as a tool for motivating workers (Bowey, 1997;Dzimbiri, 2009;White, 1973).This school of thought is backed up by contingency theory, which asserts that reward preferences vary from individual to individual depending on individual differences and needs as well as prevailing forces in the wider environment (Blackburn & Mann, 1979;Bowey, 1997;Burns & Stalker, 1962;Woodward, 1958).Franco et al.'s (2004)  The study focuses on assessing the motivational value of rewards amongst health professionals in Malawi's MoH in an effort to support the development of effective systems, practices and strategies that will improve the motivational value of rewards for the ministry's health professionals.It is theoretically structured around the above reward-related problems (refer to the whole 'Problem statement'), objectives of a reward system and schools of thought (in the 'rewardmotivation' debate) presented in this section.

Method Research approach
The study was an explanatory and evaluative hybrid of quantitative and qualitative research approaches.The explanatory approach, which was quantitative in nature, covered the development and testing of hypotheses to explain the relationship amongst variables expressed under Sections B2 to E of the data collection instrument.This approach helped to scientifically investigate and explain the reward-related problems faced by health professionals (Babbie, 2004(Babbie, , 2005(Babbie, , 2007;;Babbie & Mouton, 2001;Collis & Hussey, 2003;Berry, 2003;Burns, 2000;Clarke, 1995;Cresswell, 1994Cresswell, , 2009;;DSS Research, 2001).The evaluative approach, on the other hand, was qualitative in nature, and covered a literature review, interviews, focus group discussions and administration of Section F of the data collection instrument, This approach helped to gain insights into and familiarity with the current 'reward-motivation' debate in the researcher's quest to appraise the application of models and theories of rewards and motivation in Malawi's MoH (Ader & Mellenbergh, 2008;Decker, 1997;David & Sutton, 2004;Alston & Bowles, 1998).

Target population
For this study, the population (sampling frame) consisted of all health professionals in Malawi's MoH.

Sampling
Ten percent (10%) of the health professionals (i.e.571 of 5707 of the ministry's health professionals) were systematically selected and became participants in the study.All (100%) of these research participants completed and returned copies of the self-administered questionnaire.All elements of the population had equal representation in the frame.

Data collection instrument
A structured questionnaire was developed using the data collected from the desk research or literature review, focus group discussions and interviews.
The questionnaire had six sections, as follows: • Section A contained biographic information of the respondents such as gender, age group, qualifications, registration with professional institutions, status of employment, and length of service (number of years working) with the MoH (refer to Table 1 for the rating or scoring of each sub-item).• Section B1 aimed at assessing the respondents' knowledge of the reward system(s) (time-based or basic rate or performance-based or competence-based) operational in Malawi's MoH.Respondents were required to indicate 'Yes' (with a score of 1), 'No' (with a score of 2) or 'Don't know' (with a score of 3) respectively, implying that a reward system was operational, not operational or that it was not known whether or not it was operational in Malawi's MoH.• Section B2 aimed at assessing the knowledge and attitudes of respondents regarding the objectives of the reward system(s) operational in Malawi's MoH -whether it is able to attract the right health professionals for the right positions, retain the right health professionals for the right positions, motivate the right health professionals for the right positions, comply with the existing labour laws and policies, and compensate health professionals for services rendered.• Section C aimed at assessing the opinions of the respondents regarding the motivational value of financial and non-financial rewards in Malawi's MoH.Specific issues or items to be assessed were as follows: whether or not the respondents were primarily motivated by financial or non-financial rewards; whether or not the respondents essentially worked for financial or non-financial rewards; whether or not financial or non-financial rewards were adequate; whether or not financial or non-financial rewards were linked to performance; and whether or not current financial or non-financial rewards are able to drive the Ministry's productivity.For Sections B2, C, D and E of the questionnaire, statements that were used to measure the factors were scored on a five-point Likert-type scale.Thus, the scores for the item responses were on a scale of 1 to 5 as follows: strongly agree (with a score of 1), agree (with a score of 2), neutral (with a score of 3), agree (with a score of 4) and strongly disagree (with a score of 5).The questionnaire had not been used before.Pilot-studies were therefore conducted (followed by adjustments to the questionnaire) to improve its validity and reliability before administering it to the systematically sampled participants.

Data collection procedure
Empirical data were collected through the use a selfadministered questionnaire, focus group discussions and interviews.The questionnaire was composed using the data collected from desk research (literature review), focus group discussions and interviews.

Data analysis
Descriptive and inferential statistics were used to analyse quantitative data from the questionnaire.Such data were first coded and summarised before being entered into the computer using the Statistical Package for Social Science (SPSS) package (version 16).The package was chosen because of its convenience in analysing and interpreting data (Barry, 1998;Brace, Kemp & Snelgar, 2000;Bryman & Cramer, 1999;George & Mallery, 2003;Salkind, 2000).
The statistical techniques used to analyse the data were item and factor analyses.Such analyses helped to determine whether or not items loaded significantly on the factors they were measuring.The analyses further helped to determine the internal consistency levels of the summated factor scores through the calculation of Cronbach's coefficient alphas as displayed in the 'results' heading below.The SEM technique was also used to test and estimate the causal relations amongst variables and determine the model fit for the hypothesised and re-specified models (Bentler, 1980(Bentler, , 1990;;Hooper, 2008;Steven, 1991;Stevens, 1996).

Development of hypothesised model
The relationships between factors were measured through a hypothesised model (a collection of logical relationships amongst variables or factors derived from theories and concepts) (Bentler, 1980(Bentler, , 1990;;Steven, 1991;Stevens, 1996).
Before testing, the model envisaged that objectives of a reward system have a significant mediating effect on the rewards and reward-related satisfaction of health professionals of MPHS.It further suggested that there is a good fit between exogenous or independent variables (financial and nonfinancial rewards) and endogenous or dependent variables (reward-related problems) with 'objectives of a reward system' as a mediating variable.The relationships amongst the variables in question are displayed in Figure 1.
In this context, an exogenous or independent variable refers to a factor or construct whose values (when manipulated by the researcher) creates effects on a dependent variable; a mediating variable refers to an intervening or intermediary factor or construct which is part of the causal chain; and an endogenous or dependent variable refers to a factor  or construct whose value should be observed when an independent variable changes and represents a response, behaviour or outcome the researcher wishes to explain or predict (Gorsuch, 1983;Kim, 1995;Kim & Mueller, 1978;Losh, 2002;Ratner, 2011;Routio, 2007).The hypothesised model, which was developed from theory (Armstrong, 1993(Armstrong, , 1999;;Bratton & Gold, 2007;Koala Consulting and Training, 2008;Torrington, Hall & Taylor, 2005), was statistically tested in the 'results' heading below.

Confirming reliability and validity of the questionnaire
Reliability and validity of the questionnaire (research instrument) were confirmed as indicated in the following sub-sections.

Trustworthiness Reliability
An item analysis was conducted in order to assess the reliability of the scores derived from the research instrument.Cronbach's coefficient alphas were calculated to determine the internal consistency of the instrument.All factors reported excellent reliability statistics.Cronbach's alphas, which ranged from 0.96 to 0.99, exceeded the recommended threshold level of 0.70 for acceptable reliability, as displayed in Table 1.
Thus the item analysis results indicated that all factors had significant Cronbach's alpha values, suggesting very strong reliability and implying that they were all reliable for use in the study (Gliem & Gliem, 2002).The internal consistency of the instrument was also tested and confirmed during the pilot studies at the following Malawian health facilities:

Validity
To determine the validity of the instrument, a factor analysis with an oblique varimax rotation was conducted.The results showed a cumulated proportion of 90.30% which confirmed the validity of the questionnaire for use in this study.The item loadings for each factor were as follows: • Factor 1: Items B2.1 (0.949), B2.2 (0.924), B2.3 (0.931), B2.4 (0.919) and B2.5 (0.944).This factor confirms a reward system's objectives of attracting, retaining and motivating workers; complying with the existing labour laws and policies; and compensating workers for the labour services they render to an organisation.• Factor 2a: Items C1 (0.954), C2 (0.933) and C3 (0.970).This factor confirms the use of financial rewards as motivators in organisations.
Some authors (Fabrigar, Wegener, MacCallum & Straham, 1999;Wikipedia, 2012) recommend a loading of 0.7 or higher to confirm that an item represents a particular factor in a research instrument.The above results indicate high loadings (of equal to or greater than 0.822) for all items making up each factor confirming validity of the whole research instrument.
Descriptive statistics for biographical characteristics of the respondents.
Five hundred and seventy-one (i.e., n = 571) health professionals sampled for the study completed the questionnaire.The figure represents 100% of the selected study sample.For this reason, 'n = 571' for every finding in this study.Therefore, 'n = 571' will not be specified when presenting every finding to avoid unnecessary repetition.Descriptive statistics for biographical data included gender, age, qualifications, registration, status of employment and length of service, as displayed in Table 2.
In Table 2 The respondents were requested to indicate their status of employment in Malawi's MoH (whether they were permanent, temporary, probationary or contract employees).

Statistical analysis
Statistical analysis was carried out to test the hypothesised and re-specified models as discussed in the following subsections.

Hypothesis testing
One of the hypotheses formulated and tested in this study was that objectives of a reward system have a significant mediating effect on rewards and reward-related satisfaction of health professionals of MPHS.This hypothesis suggests that there is a good fit between financial rewards, nonfinancial rewards and reward-related problems (represented by attraction, motivation and retention; non-corrupt practices; equity of rewards; financial coping strategies; and industrial democracy) when mediated by the reward system's objectives.
In terms of theory (Armstrong, 1993(Armstrong, , 1999;;Bratton & Gold, 2007;Koala Consulting and Training, 2008;Torrington et al., 2005), the objectives of a reward system mediate or strengthen the motivational value of financial and non-financial rewards to counter reward-related problems (mentioned in the 'problem statement and research objectives' section).In this study, the researcher's proposed model (refer to Figure 1) was based on theory and tested by means of the SEM technique in consonance with former research publications (Bentler, 1980(Bentler, , 1990;;Fan Thompson & Wang, 1999;Hooper, 2008;Steven, 1991;Stevens, 1996).However, the analysis of the model (through the SEM exercise) collapsed suggesting that the sample in this study does not support the measurement model.It could be interpreted that the objectives of a reward system do not have a mediating effect in addressing reward-related problems through financial and non-financial rewards in Malawi's MoH.The hypothesis was therefore rejected.Consequently, the mediating factor (objectives of a reward system) was omitted when developing and testing a re-specified model, presented below.

The re-specified model
Some researchers (Bagozzi, 2012;Goldberger, 1972;Pearl, 2000) assert that whenever necessary, a model must be modified (re-specified) to improve the fit and estimate the most likely relationships between variables.Modifications that improve model fit should be treated as potential changes made to the model.In addition to improvements in the model fit, the modifications (re-specifications) should also make theoretical sense.Hooper (2008) and Kenny (2011) agree that when a model has been re-specified, any modifications made and correlations (causal relations) between variables must be reported.It is against the professional advice of some researchers (Bagozzi, 2012;Goldberger, 1972;Hooper, 2008;Kenny, 2011;Kerliger & Lee, 2000;Pearl, 2000) that the re-specified model (and correlations between the variables) was reported in this study.Objectives of the reward system were omitted from the re-specified model and consequently the SEM exercise produced significant results from the respecified model with several explanations.Figure 2 displays the re-specified model followed by explanations of the causal relationships amongst the variables.
In Figure 2, it is clear that there are positive causal relationships (correlations) between financial rewards and reward-related problems faced by health professionals in Malawi's MoH.The weakest correlation exists between financial rewards and erosion of industrial democracy.The strongest association (with a coefficient of 0.46) exists between financial rewards and failure to attract, motivate and retain health professionals followed by the correlation between financial rewards and perception of inequity of rewards (with a value of 0.45).The value of the correlation between financial rewards and coping strategies to supplement financial rewards is 0.44, whilst the value of the correlation between financial rewards and engagement in corrupt practices is 0.26.The results confirm that there is a positive relationship between financial rewards and alleged reward-related problems being faced by health professionals in Malawi's MoH.These results are statistically significant (p < 0.05) (Hooper, 2008) and imply the following: • Inadequate financial rewards have resulted in the failure of Malawi's MoH to attract, motivate and retain staff.• There is a perception of inequity of rewards amongst health professionals in Malawi's MoH.• Health professionals develop and use coping strategies to supplement their financial rewards.• Health professionals engage in corrupt practices to supplement their financial rewards.• There is an erosion of industrial democracy in Malawi's MoH due to inadequate financial rewards.
Qualitative responses (from both male and female respondents, married and not married within the '20 to 49 years' age range) captured under Section F of the questionnaire were reported verbatim as follows: • 'Brain drain will continue to torment Malawi's MoH if the Malawi government is not serious in managing the HR crisis created by inadequate rewards provided to health professionals'.• 'Health professionals will continue to take up job opportunities in the global market as demand for their expertise rises in high income areas'.
Qualitative responses captured during interviews and focus group discussions were coded, paraphrased and summarised as follows: • If financial rewards for health professionals are not market-driven, policy-makers should not be surprised with the massive exodus of health professionals from Malawi's Ministry of Health.• Health professionals find it frustrating to receive lower financial reward packages than expatriate technical support staff with the same or lesser qualification and experience.
The quantitative findings (supported by the qualitative responses) confirm the motivational value of financial rewards, which is an academic departure from Herzberg's (1960Herzberg's ( , 1966) ) two-factor theory which regards financial rewards as a hygiene factor and not a motivator.
Various researchers (Bentler, 1980(Bentler, , 1990;;Hooper, 2008;Steven, 1991;Stevens, 1996) advise that when selecting model fit statistics to report, the researcher should simply report the statistics that estimate the best fit.Other researchers (Fan, Thompson & Wang, 1999;Kerlinger & Lee, 2000) caution that several fit statistics must be consulted when evaluating a model fit because fit indices were developed for different rationales.No single fit meets all expectations of an ideal index.In consonance with this advice, a factor analysis was conducted to determine the model fit.Consequently, the statistical measures that estimated the reasonable fit between the proposed model and the empirical data were a normed fit index (NFI), (Bentler & Bonnet, 1980) with a value of 0.86; a non-normed fit index (NNFI) (Bentler & Bonnet, 1980) with a value of 0.86; and a comparative fit index (CFI), (Bentler, 1990) with the value of 0.88.
Deducing from Figure 2, this researcher has established an inverse relationship (negative correlation) between financial rewards and non-financial rewards as graphically demonstrated in Figure 3.
In the graph presented in Figure 3, the horizontal axis denotes financial rewards (FR) whilst the vertical axis denotes nonfinancial rewards (NFR).The NFR-FR line or curve represents the relationship between non-financial rewards and financial rewards.
Theoretically, the graphic model shows that, ceteris paribus, health professionals with initial financial rewards at FR 0 will demand initial non-financial rewards at NFR 0 from the employer (i.e., Malawi's MoH).If financial rewards are increased by the employer to FR 1 , demand for non-financial rewards by health professionals will decrease to NFR 1 .If financial rewards are decreased by the employer to FR 2 , demand for non-financial rewards by health professionals will increase to NFR 2 .Decreasing financial rewards to 0 1 by the employer implies the end of the employment contract and demand for non-financial rewards by health professionals disappears.On the other hand, increasing financial rewards to 0 2 will result in zero demand for non-financial rewards by health professionals.
This researcher is of the opinion that the graphic model displayed in Figure 3   financial implications.These are housing or accommodation or shelter, training opportunities, development opportunities, promotional opportunities, company car, good working environment, club membership and annual award ceremonies to recognise employees' contributions (Armstrong & Mullins, 2007;Carrel et al., 1998;Howard, 1997;Michalowicz & Male, 2011).This researcher argues that these are regarded as non-financial rewards due to lack of better terminology to describe them.Ceteris paribus, demand for so-called nonfinancial rewards by the workers is likely to increase if they do not get adequate financial rewards from their employers.
The employer cannot provide these non-financial rewards to the workers free of charge.It is therefore an economic blunder to think that by providing them to the workers, the employer is avoiding financial implications.If workers get adequate financial rewards from their employers, they can get these so-called non-financial rewards on their own thereby increasing intrinsic motivation.For instance, they can acquire their own accommodation or housing, train and develop themselves and create their own good working environment using financial rewards they get from their respective employers.
This researcher concludes that both financial and nonfinancial rewards are important in the world of work as demonstrated by the graphic model displayed in Figure 3.However, the success of an organisation in maintaining the motivational value of both financial and non-financial rewards (whilst improving its performance or productivity) therefore depends on its ability to balance the categories of rewards, including those with no financial implications.These are praise, delegation of work, job enrichment, job enlargement, participation in decision making, appreciation for personal contribution, giving employees a new or improved job title, celebrating the anniversary of their joining the company and a handwritten 'thank you' note (Michalowicz & Male, 2011).

Discussion
The aim of the study was to assess the motivational value of rewards amongst health professionals in Malawi's MoH.So far, literature reveals that no consensus has been reached regarding the motivational value of financial and non-financial rewards.As a consequence, organisations are applying theories and models of motivation selectively depending on their beliefs, ideological framework of values and assumptions of variations in the relative importance of rewards at their disposal (Howard, 1997;Torrington et al., 2005).
In this study, the hypothesised model was based on theory and tested by means of correlations through the use of the SEM technique.As explained above, the analysis deleted the 'objectives of a reward system' as a mediating factor.As a consequence, an initial fit of the empirical data on the proposed model was not established.These findings suggest that the objectives of a reward system do not have a mediating effect on the reduction of reward-related problems.
The re-specified model (Figure 2) was also tested using the SEM technique and a reasonable fit was established between the re-specified model and empirical data.The statistical measures that estimated the reasonable fit were NFI, NNFI and CFI.The findings of testing the re-specified model suggest that there are positive causal relationships between financial rewards and reward-related problems faced by health professionals in Malawi's MoH.These findings imply that if the employer increases financial rewards for health professionals, the probability of eliminating reward-related problems increases as well, and vice versa.
The findings displayed in the graphical model (refer to Figure 3) suggest that there is an inverse relationship between financial and non-financial rewards.(1960,1966) downplay the use of financial rewards as motivators, they are critical to work motivation especially in the developing countries where the struggle for basic needs (in order to survive) is a reality (Caudron, 1993;Robbins, 2010).These findings do not suggest that non-financial rewards should be completely ignored.They imply that arguments or criticisms against rewards (whether financial or non-financial) should be made with caution because reward preferences vary from individual to individual depending on individual differences and needs as well as the prevailing environmental forces or factors, as suggested by Franco et al.'s (2004) model and supported by the contingency theory of motivation (Blackburn & Mann, 1979;Bowey, 1997;Landy & Conte, 2010;White, 1973).
Therefore, this study is, to the researcher's knowledge, one of the first studies to assess the motivational value of rewards amongst health professionals in the public health sector.It is unique in the sense that its findings indicate that both financial and non-financial rewards are indispensable in employment relationships.Furthermore, in the world of work, no relationship exists between rewards, objectives of a reward system and reward-related problems, as manifested by the collapsing of the hypothesised model when an attempt was made to statistically test it.

Managerial implications and recommendations
The findings of the hypothesised model suggest to managers and/or policy-makers that the objectives of a reward system have no mediating effect between the rewards and rewardrelated problems faced by health professionals in Malawi's MoH.The finding derived from both re-specified and graphical models symbolise a pragmatic departure from theoretical claims that financial rewards (salaries or money in general) are not a motivator (Herzberg, 1960(Herzberg, , 1966)).Therefore, although Franco et al.'s (2004) (Amar, 1994;De Bruyne, 2001;Kapeleta, 2002).Managers and policymakers should always bear in mind that the commitment of health professionals to improving and maintaining quality standards of health care cannot be taken for granted.Offering competitive rewards and ensuring that reward systems and practices are efficient, reasonable, fair, consistent and acceptable to all health professionals is a positive step to attracting and retaining them as well as winning their commitment (Amar, 1994).

Limitations of the study
The major limitation pertaining to the literature was that there is a massive body of contemporary writing on motivation and rewards.The preferences of the researcher determined the choice of the key texts which, in his view, were relevant to the core theme of the study.For empirical research, behaviour of the respondents, who were systematically sampled for the study, was not directly assessed because the researcher relied solely on the self-administered questionnaire to collect data from them.Consequently, the researcher has failed to articulate the behaviour and attitudes of the respondents during the data collection exercise.

Recommendations for future research
The first recommendation for future research relates to the inverse (causal) relationship that has been established between financial and non-financial rewards as demonstrated in a graphic model (Figure 3).The graphical model demonstrates that, ceteris paribus, if the employer increases financial rewards, demand for non-financial rewards by workers decreases and vice versa.The researcher is of the view that this model should not be taken lightly by other researchers, scholars, consultants and practitioners since it is not backed up by theory that is mostly founded on Western studies.Further empirical research is recommended to re-test the model in the public health sectors of developing countries with other samples.
The second recommendation relates to the view that financial rewards are of primary importance (compared to non-financial rewards) in any employment relationship in the developing countries since the reciprocity principle of the contract of employment states that workers are entitled to financial rewards when they have placed their labour services at the disposal of an employer (Basson et al., 2002;Finnemore, 2002).Presently there is a pool of empirical research findings that organisations can succeed in motivating their workers and improving their productivity through the use of financial rewards only.However, no empirical evidence is readily available supporting the argument that through the use of only non-financial rewards, organisations can succeed in motivating their workers and improving their productivity.Plainly speaking, the motivational value of non-financial rewards has been praised in the context where financial rewards have not been completely ignored (Dewhurst, Guthridge & Mohr, 2009;Silverman, 2004).It is therefore recommended further that scientifically informed research should be conducted to bring to light examples of wellestablished business organisations in developing countries whose workers have only non-financial reward packages governing their contracts of employment.
A comparative analysis of performance of the two categories of business organisations (that is those whose workers only have financial reward packages governing their contracts of employment and those whose workers only have nonfinancial reward packages governing their contracts of employment) should also be done and reported on to end the debate.
The third and final recommendation relates to this researcher's belief that a candle does not lose anything by lighting another candle.The researcher is convinced that 'praise' or 'thank you' as a form of recognition, positive feedback, participative decision-making, job autonomy, delegation, job enlargement and job enrichment are unquestionably non-financial rewards.However, inspired scientific clarifications should be made (through scientifically informed research) by the proponents of non-financial rewards on how other alleged non-financial rewards (especially career advancement, housing or accommodation or shelter, training and development promotional opportunities, company car, good working environment, club membership, and annual award ceremonies to recognise employees' contributions) can be provided by an employer without financial implications attached to them.Otherwise the researcher is of the view that these non-financial rewards are basically financial rewards offered to health professionals (or any workers) in a disguised form.

Conclusion
From the above discussions and debates, it is clear that a lot has been said, written and talked about regarding the motivational value of rewards in the field of industrial and organisational psychology.What is indisputable is that rewards are indispensable in every employment relationship.Offering competitive rewards and ensuring that reward systems and practices are efficient, reasonable, fair, consistent and acceptable to all health professionals is a positive step to attracting and retaining them, and ultimately winning their commitment (Amar, 1994;Dieleman, Cuong, Anh & Martineau, 2003;Kapeleta, 2002).

FIGURE 1 :
FIGURE 1: Hypothesised model of reward categories and their consequences.
is a true reflection of what happens in the real world of work especially in developing countries where the frequently cited non-financial rewards have

FIGURE 2 :FIGURE 3 :
FIGURE 2:The re-specified model displaying corrections between variables.

•
To investigate the perceptions of a sample of health professionals about the efficiency and motivational value of the financial rewards available in Malawi's MoH.To suggest a model that will help to bridge the gaps in the existing theories of rewards and motivation.
According to Mr Hillary Chimota, Controller of Human Resource Management and Development in Malawi's MoH (personal communication, 09 June 2011), 5707 out of 33 376 health workers in the MPHS are health professionals on the ministry's payroll.These 5707 health professionals formed the population for this study because they belong to the core function or business of Malawi's MoH whilst the rest of the workers (27 669) are support staff belonging to the common service (support function) of the whole Malawi civil service.
claim that with the complement of qualitative techniques, descriptive and inferential research techniques are useful in identifying problems with current practices and making judgements.Such techniques are also useful in determining what others in similar situations are doing.The findings of a study that has utilised quantitative inferential techniques are generalisable and replicable to the whole target population.

TABLE 1 :
Item analysis results.

TABLE 2 :
Biographical characteristics of the respondents.
Franco et al. (2004)wards are increased by the employer, demand for non-financial rewards decreases and vice versa.The rationale behind this inverse relationship is that in the real world of work, ceteris paribus, workers use financial rewards (if they are adequate) to get their non-financial needs or requirements.If financial rewards are not adequate, demand for non-financial needs or requirements from the employer increases.These findings imply that both financial and non-financial rewards are very important in any employment relationship and signify a paradigm shift in the way the motivational value of rewards has been viewed in the field of industrial psychology.Both the re-specified and graphical models symbolise a pragmatic departure from the model ofFranco et al. (2004), which leans towards the use of non-financial rewards and suggests that financial rewards should be used with caution.
model proposes that financial rewards should be used with caution, this researcher's opinion is that statements about rewards should be made with caution because reward preferences differ from worker to worker amidst various forces operating in the world of work and society as a whole.Rewards (as well as their related practices, systems and strategies) are the major cause of dissatisfaction (and resultantly brain drain) amongst the health professionals in Malawi's MoH because they are perceived by them to be inequitable, inadequate and unfair.The inverse relationship between financial and non-financial rewards that has been graphically demonstrated in Figure3is an empirical indication that both financial and non-financial rewards are indispensable in every employment relationship.They are a powerful means of informing workers what is actually valued or not valued in the organisation