Quantifying the Impact of Work Environment Factors on the Variability of Labour Productivity in Wall Plastering

Variability in labour productivity is a performance inhibitor and a determinant of effective and ineffective projects. It has hampered the intercomparison of construction projects and the accurate forecasting of project duration and cost. This study chose wall plastering activities as a case study. This study aims to quantify the impact of work environment factors on the variability of labour productivity. Data were collected using direct site observations and structured questionnaires. The results revealed that waiting for materials (62.4%), being on the job but not working (52.6%), and work area congestion (52.5%) all had negative effects on labour productivity variance. Other negative factors include rework (51.7%), waiting for tools/equipment (51.1%), waiting for information (47.2%) and weather changes. The overall average daily productivity was 1.268 whr/m 2 , baseline productivity = 0.993 whr/m 2 and variation in daily productivity = 22.08%. The findings identified significant work environment factors and quantified their impacts on labour productivity variability in plastering activity. The results indicate that work environment factors during work in progress significantly impact the variability of labour productivity in plastering work, and ample consideration should be given to its effects.


Introduction
In a competitive business environment such as the construction industry, improving the labour productivity of the construction workforce is crucial to the survival of any construction firm, as labour costs typically account for 30% to 50% of the project's total cost. (Jakas & Bita, 2012).
Labour productivity is often estimated and priced based on the time required to accomplish each project component. During on-site production, work environment factors come into the picture, interfering with or disrupting work progress. As a result, fluctuations in daily labour productivity occurs. Labour productivity variability is the difference in daily, weekly, or monthly labour output within and among gangs . It is an inhibitor of performance, and it determines effective and ineffective projects.
Variability induces impact and unexpected conditions, making project goals unstable and obscuring the means of achieving them (Gerek et al., 2016).
Few studies have investigated the effects of the foregoing factors on labour productivity variability in specific work environments, trades/crafts, and projects. However, most of these research adopted questionnaires and activity sampling methods to collect data, but they lacked information on the craftsmens' productive time. (Shashank & Hazra, 2014;Talhouni, 1990). Thus, this study seeks to identify work environment factors causing variability of labour productivity in wall plastering, measure labour productivity output using time study, and qualifying the impact of the work environment factors identified.

Labour Productivity
Previous studies have established that no universally accepted definition of productivity exists. (Hanna et al., 2008;Swapnil & Biswas, 2015;Gerek et al., 2016;Rao & Sudhanva,2017). Various definitions have been proposed based on the measurement method, the measurement or study's objective, and the end-users of the data collected from the measurement (Agbo, 2014).
In these definitions, productivity is viewed as a measure of the outputs obtained due to the combination of inputs (Rao & Sudhanva, 2017;Agbo & Izam, 2019). Based on this standpoint, two general measures of productivity Labour Productivity Variability in Wall Plastering were considered: total factor productivity (TFP) and partial factor productivity (PFP) (Gerek et al., 2016). Total factor productivity is the productivity calculated when all inputs are accounted for, whether tangible or intangible (Sweis et al., 2009). The TFP is used to optimise the resource inputs required to produce the desired outcome. The TFP is calculated as follows: Partial factor productivity (PFP) or unit rate productivity, on the other hand, is frequently referred to as labour productivity. It aims to build a connection between outputs and a subset of inputs. When focusing on the effectiveness and efficiency of a small number of input resources, the PFP becomes the most suitable method for measuring productivity. (Russel & Taylor, 2009).
Most contractors prefer to use equation 3, which is the inverse of equation 2.
The reason is that most contractors are more concerned with the number of hours a worker works every day because they pay their workers by the hour. (Gopal & Murali, 2015;Odesola et al., 2015). Hence, this study adopts equation 3 because 65% of the construction firms pay their workers by calculating the number of hours worked out of 8 working hours per day.

Productivity Measurement
Productivity measurement is a performance indicator that measures the efficiency of a construction firm to current input resources such as labour, materials, equipment, among others. (Ali, Smith, & Chion, 2010;Chan & Kaka, 2010). Several techniques of productivity measurement are available. The choice of method(s) depends on the purpose of the research, the type of data required, and the resources available (Swpnil & Biswas, 2015). Noor (1992) grouped these methods into continuous observation and intermittent observation. One way to minimise the influence of factors affecting labour productivity output on construction sites is by obtaining quantitative site data on factors affecting labour productivity using appropriate productivity measurement methods on site. The data gathered via accurate on-site measuring methods can be utilised to model productivity loss and its impact. (Chan & Kaka (2010). Olomolaiye and Ogunlana (1987) evaluated the productivity of building artisans in wall plastering in Lagos, measured their labour output through work sampling, and reported that the average daily productivity was 9.3m 2 , based on eight work hours per day. However, the author's findings did not indicate any variability.
Similarly, Odesola et al. (2015) investigated labour productivity in wall plastering in six states in southern Nigeria, using direct continuous observation on the site. According to the study's findings, the average daily labour productivity was 2.68m 2 /hr. Additionally, significant variation exists within and among gangs and the various projects studied. Similarly, Udegbe (2005) Labour Productivity Variability in Wall Plastering examined labour force output in the plastering industry in Edo state, Nigeria, and discovered an average daily labour output of 16.65m 2 .

Labour Productivity Variability
Labour productivity variability is the differences in daily, weekly, or monthly labour output or labour productivity within and among gangs (Swapnil & Biswas, 2015). It is a well-established fact that labour productivity fluctuates throughout an activity (Anu & Sudhakumar, 2013). The variance results from the existence of work environment factors during the execution of the activity, which can be classified as management factors such as a lack of material, overcrowding in workspaces, rework, and gang composition.
Additionally, technical factors such as incomplete design, inexperience, supervisor incompetence, and individual factors such as skill level differences, fatigue, and other behavioural factors that influence workers' performance, also contribute to the output rate variation observed in practice. (El-Rays & Moselhi, 2001;Abdel-Razek et al., 2006).
Factors affecting productivity vary from gang to gang, site to site, project to project, and from day to day (Talhouni, 1990;Rao & Sudhanva, 2017).
According to Idiake (2014), artisans' output variations on construction sites are caused by a lack of experience, competence, and overtime work lasting more than 30 minutes. He calculated that wall plastering generated a variance of 39.56%. Song and Abou-Rizk (2008) stated that design, management, working hours, congestion, and weather would increase variability in performance and make productivity comparison almost impossible. The cumulative effect of these various factors causes random and systematic disturbances to performance in performance-intensive operations (Mohammed & Moselhi, 2005). If the effect of these factors is discounted from the actual performance, there will be a smooth and non-variable performance curve.
According to Ofori-Kuragu, Baiden, and Edum-Fotuse (2010), productivity variation among individual artisans was 4.1% (60%) on selected housing sites with a similar design. Similarly, Talhouni (1990) discovered that the variability in bricklaying and plastering production was 2:3.1 (+39 percent) and that a bricklayer gang could produce twice the average output. However, the investigation took no consideration of design differences and interference factors. The author noted considerable variability on-site, particularly on construction sites where gangs worked.

Causes of Labour Productivity Variability
Faridi and El-Sayegh (2006) developed a framework for a labour-intensive forecasting model on construction sites. They outlined factors causing variation in productivity into project-specific, project dependent, and regiondependent. These factors can be classified as physical and non-physical characteristics of projects that affect construction labour productivity.
Project size, location, building height, the degree to which engineering overlaps construction, and project administration by a strong project management team are the physical and non-physical characteristics described by Enshassi et al. (2007). In a similar vein, Thomas and Horman (2006) listed the significant causes of variability in brick/block layers' productivity as site delays, variation in length working days, and gang composition.
Frimpong, Oluwoye, and Crawford (2003) defined site delay as a situation where the workforce is either stopped from working or is functioning inefficiently. This situation usually arises during a project's construction phase.
Delays have a substantial impact on construction employees' productivity.
Their effect varies depending on the type of construction delay(s) that occur (Hegab & Smith, 2007).
Construction site delays have been described in a variety of ways by different researchers. Hegab and Smith (2007), for example, distinguished between intrinsic and extrinsic delays. Intrinsic delays occur due to a trade's operational characteristics or the features of a particular construction site.
Examples include waiting for the scaffolding installation and delivering materials from the stockpile to the construction site. On the other hand, extrinsic delays are induced by nature and over which management has no control, such as inclement weather and natural disasters. Thomas and Horman (2006) classified the significant causes of construction site delay as weather, material, plant/tools, sequence, rework and instructions. Hanna, Taylor, and Sullivan (2005) examined the effects of longer working hours on workers' output. They concluded that increasing the number of hours worked per day per week would decrease productivity.
According to Goodrum, Zhai, and Yasin (2009), increasing the length of the working day to 12 hours or more each week will result in a 55% decrease in production. Naturally, when workers are required to work longer than their regular working hours per day or week, fatigue sets in, resulting in a diminishing return on their output. (i.e., decreased production).
Hanna, Chan, Lackney, and Sullivan (2007) examined the man-hours necessary to construct individual housing units in Ireland. While appraising the block laying process, they discovered that the optimal gang size for manhours required per individual block was two block layers and one labourer.
Their investigation revealed that the larger the gang size, the more man-hours per house that are required. However, it is crucial to note that Hanna et al.
(2007) recommend a gang size of 2:1 (block layer to the worker) for walls less than 1.4 metres in height.

Research Method
This study employed a mixed strategy to collect data in order to meet the research objectives. This technique involved two types of data: quantitative data collected using a structured questionnaire and qualitative data obtained by direct continuous site observation on-site.

Population and Population Sampling
The study population is any group of persons, objects, or institutions that exhibit one or more of the research characteristics (Bernold & Milton, 2010).
The study population consists of all registered Estate Developers in Abuja who currently have on-site projects. The Federal Capital Development Authority (FCDA) provided a list of registered estate developers, which shows only 27 registered contractors with ongoing projects in Abuja. These 27 contractors comprised the study population, from which the sample size was calculated.
The following formula was used to determine the sample size for this study: Data Collection and Analysis.
A time study was used to collect data using questionnaires and direct site observation. Utilising these two methods became necessary because the data collected by questionnaire was intended to supplement the data gathered through direct site observation. A structured questionnaire was designed to elicit information about the factors affecting plasterers' labour productivity from project managers, engineers, supervisors, and plaster masons. Eight questionnaires were distributed at some sites, while seven were administered at others, depending on the number of respondents willing to reply to the questions. In total, 190 questionnaires were distributed, with 152 duly completed and returned.
The information from the questionnaires was analysed using the relative importance index (RII). RII = 5n5 + 4n4 +3n3 + 2n2 +n1/5(n1 +n2 +n3 + n4 + n5) x 100. --------- (6) In direct continuous observation, the first step was to identify 25 ongoing public building projects in the study area which use a standard sandcrete block of 225mm x 225mm x 450mm, a prototype, and all plastering work were on the ground floor. The researcher and his observers were then granted free access to the site after receiving official clearance from the client and contractors. Before the commencement of the study, the workers to be observed were assembled on-site, and the purpose of the observation was explained to them to avoid the "Hawthorn" effect, that is, workers working diligently because they are being observed. To avoid bias, when a specific gang is selected, it is kept hidden that they are being watched while on-site research assistants make observations. An average of 28 observations was made on each of the sites during the 30 days observation period.
Each day, the research assistants arrived 20 minutes before the start of work.
Coefficient of productivity variation(CPV) = PV X100/Baseline productivity------- 5. Results and Discussion. Table 1 shows the results of productivity measurement in wall plastering in Abuja, Nigeria. The expected average daily output quantities and daily productivity as shown in the table represent the minimum standard of daily output quantities and daily productivity in wall plastering for a gang of two members: a mason and a labourer, respectively, 22 m2 and 0.727 whr/m 2 . This information was gathered from site engineers, supervisors, and foremen who were directly responsible for monitoring the day-today labour output and productivity of construction craftsmen on site, particularly those on the monthly payroll of the contracting company. The average daily labour output and productivity, according to this study, were 17.33m 2 and 0.916whr/m 2 , respectively. These results were slightly less than the minimal level established in Abuja for wall plastering. On the contrary, the results indicated that a few individual projects' average daily labour output and labour productivity (projects 18, 19, 20, and 24) exceeded the Abuja minimum standard for wall plastering. This analysis suggests that the project managers of better-performing projects had better managerial abilities than the poorly performing projects. When a project is appropriately managed, it results in increased labour production, improved performance, and low variability; when managed poorly, it results in decreased labour output, poor performance, and high variability (Gerek et al., 2016).

Productivities and Output Quantities in Wall Plastering
According to this study, project 19 was the best managed and performed due to its low variability and higher labour productivity. Similarly, this research revealed that project 004 was the worst performing and managed project due to its high variability and low labour production. The overall and baseline productivity averages were 0.808whr/m 2 and 0.712whr/m 2 , respectively. The trends of these productivities follow average daily productivity and daily labour output. The labour productivity and baseline productivity trend are depicted in Figures 1 and 2 for projects 19 and 004. The baseline productivity was computed using the following formula: Baseline productivity = summation of work hrs/output quantity in n workdays Idiake (2014) conducted a similar analysis on wall plastering in Abuja and discovered that the average daily productivity of masons varies between 0.753whr/m 2 and 1.415whr/m 2 . Similarly, Swapnil and Biswas (2015) examined labour productivity in wall plastering and discovered an average daily output of 1.31m 2 /hr.
Although the average daily labour output and daily productivity determined in this study were less than the minimum standard, they were comparable to those found in previous studies. Hence, the daily average labour output of 17.33m2 and daily productivity of 0.916whr/m 2 could be used to estimate wall plastering in Abuja. Table 2 shows the result of the variability of labour productivity. The coefficient of productivity variation (CPV) was computed for all the projects investigated using the formula below. CPV = PV X 100/(Baseline productivity).

Labour Productivity Variability in Wall Plastering.
The CPV results revealed that the rate of labour productivity variability in wall plastering ranges from 8.76% to 63.3%, with project 04 having the highest percentage of variability of 63.3% and project 19 having the lowest percentage of variability of 8.76%. The overall average variability was 22.08%.
According to Anu and Sudhakumar (2013), a higher value of labour productivity variability is a sign of poor performance, which is a pointer to poor management of such projects. On the other hand, a lower value of variability of labour productivity is a sign of better performance and better management of such projects.  The findings of this study on average daily labour productivity variability were similar to those of previous studies. For instance, Swapnil and Biswas (2015) reported that the coefficient of labour productivity variability in wall plastering ranges from 35 to 147.5%. Furthermore, Idiake (2015) reported that the coefficient of labour variability in wall plastering in Abuja was 28.26%. He equally observed that the significant causes of variability of labour productivity of craftsmen on construction sites are interruption (a delay that lasted not more than 2 hours) and disruption (a delay that lasted above 2 hours). Figures 3 and 4 illustrate the trend of labour productivity variability for projects 19 and 04, respectively. The movement of the trend reveals a rise and fall pattern.

Wall Plastering
A checklist of twenty-five work environment factors drawn from previous studies was presented to respondents who worked as plasterers on the construction sites sampled. They were instructed to tick, appropriately, the factors that affect their labour output and labour productivity.
On average, 15 factors were ticked as factors affecting their labour output and labour productivity. These variables were similar to those identified in prior research (Thomas et al., 2002;2006;Vaishant & Kansal, 2016). The ranking of the variables according to their severity revealed that waiting for materials came in first place on the ranking scale, with a 60.6 importance index. This figure was followed by waiting for instruction and rework, both of which had an importance index of 57.24 and 56.69. Accident was ranked as the least influential factor by the plasterers. These findings corroborate those of Udegbe (2005).

Quantifying the impact of Work Environment Factors on Labour Productivity Variability in Plastering Activity
In order to quantify the impact of work environment factors on labour productivity variability in wall plastering activity, a multiple regression model was developed. Multiple regression allows a researcher to predict Y scores based on several X scores. Hence, the multiple regression model was used to predict the relationship (impact) between work environment factors and labour output in wall plastering. In other words, the effect of X1, X2, X3,..., X15 on the variability of Y was predicted using Y's scores.
The model is in the following structure:  accidents, which accounted for 11%, plant/equipment breakdown, which accounted for 21.0 %, and waiting for instruction, which accounted for 21.6 %.
The effect of these factors on variability was statistically insignificant and thus negligible. This means that while these factors have a minimal effect on the variability of labour productivity in wall plastering in this study, they may considerably affect other masonry construction activities such as blockwork and concrete work.

CONCLUSION.
This study investigated the impact of work environment factors on labour productivity variability in wall plastering. It was concluded that there is significant labour productivity variability in plastering activity for projects surveyed. The causes of variability were ascribed to the presence of certain work environment variables that interrupt (cause a delay of between 1and 2 hours) and disrupt (cause a delay of more than two hours) work progress. It was also observed that factors such as being on the job but not working