The respondents’ age, gender, race/ethnicity, marital status, and education level, as well as the respondents’ personal and professional histories, were the descriptive data analysis outcomes. Percentages, averages, and standard deviations were utilised as descriptive statistics.
The majority of respondents (87.2%) were male; the median age range was 26–35 (51.1%); and 4.8% of the sample was younger than 20 years old.
Most of the respondents worked as engineers (27.8%), followed by contractors (22.1%), timekeepers (22.1%), and facilitators (22.1%).
The majority of respondents (66.1%) either held a bachelor’s degree (36.3%) or a technical certification (30.6%), while 16.9% had completed high school. While nearly half of respondents (47%) had two to five years of experience in the workforce, just over half (53.1%) had six years of experience or more. This demonstrates that the respondents were capable of working in the construction business and had the necessary experience to provide data that may be used to draw conclusions on parameters measuring labour productivity. The geographic distribution of respondents was nearly even: 20.6% were from Bolgatanga, 20.6% were from Wa, 20.3% were from Tamale, 20.1% were from Kumasi, and 18.4% were from Accra. However, the majority of respondents (61.5%) were employed in the three northern regions of Ghana: Bolgatanga, Wa, and Tamale.
4.1. Results from Exploratory Factor Analysis
This part of the report details the findings from Questionnaire Section B, which aimed to identify the factors affecting the productivity of construction workers in Ghana’s most labour-intensive industry: road building. The results of the means, standard deviations, and the rank of item score of the data, as well as the exploratory factor analysis (EFA) of the results, are presented. The descriptive results revealed the ranking of all the factors from the highest to the lowest and the individual means and standard deviations of the factors.
Table 2 reveals the respondents’ rankings of the worker component (WC) that can promote the labour productivity of labour-intensive works on road construction. It shows that “the company’s incentive scheme for good performance” was ranked first, with a mean score of 4.12 and a standard deviation (SD) of 0.976; “opportunities for employees to exercise their skills” was ranked second, with a mean score of 4.11 and an SD of 0.697; “likelihood beneficiaries are paid on time” was ranked third, with a mean score of 4.10 and an SD of 0.986; “management response to settle employees’ grievances” was ranked fourth, with a mean score of 4.03 and an SD of 1.02; and “beneficiaries’ knowledge of scientific techniques” was ranked fifth, with a mean score of 3.95 and an SD of 0.839.
In addition, “beneficiaries’ attitude towards the job they have to execute” was ranked sixth, with a mean score of 3.91 and an SD of 0.852; “beneficiaries’ knowledge of career prospects” was ranked seventh, with a mean score of 3.88 and an SD of 0.921; “promotion opportunities for employees” was ranked eighth, with a mean score of 3.86 and an SD of 1.16; “employment of young beneficiaries on projects” was ranked ninth, with a mean score of 3.83 and an SD of 1.46; and “Beneficiaries’ level of experience to do their work” was ranked tenth, with a mean score of 3.77 and an SD of 0.496.
Furthermore, “likelihood older beneficiaries will be replaced by younger beneficiaries” was ranked eleventh, with a mean score of 3.69 and SD of 1.17; “level of safety achieved on projects” was ranked twelfth, with a mean score of 3.67 and an SD of 0.963; “employment of older beneficiaries from villages” was ranked thirteenth, with a mean score of 3.60 and an SD of 1.23; “employees level of awareness of company policy” was ranked fourteenth, with a mean score of 3.59 and an SD of 0.964; and “incentives used to attract young people into sector” was ranked fifteenth, with a mean score of 3.59 and an SD of 1.73.
Moreover, “the number of multi-skilled project personnel in the company” was ranked sixteenth, with a mean score of 3.57 and an SD of 0.591; “beneficiaries’ having formal training in labour-intensive works” was ranked seventeenth, with a mean score of 3.54 and an SD of 0.965; “quality of transportation facilities for beneficiaries” was ranked eighteenth, with a mean score of 3.53 and an SD of 1.057; “the usage of safety wear on-site” was ranked nineteenth, with a mean score of 3.21 and an SD of 0.961; and “degree to which safety standards on a project comply with legislated criteria” was ranked twentieth, with a mean score of 3.08 and an SD of 1.098.
4.2. Results from Exploratory Factor Analysis
The results from the EFA on the worker component that can promote the labour productivity of labour-intensive works for road construction are presented. Of the twenty (20) variables listed, the following five (5) were omitted: “workers’ having formal training in labour-intensive works” (WC1), “beneficiaries level of awareness of company policy” (WC6), “beneficiaries’ knowledge of career prospects” (WC10), “likelihood beneficiaries are paid on time” (WC12), and “degree to which safety standards on a project comply with legislated criteria”(WC15).
The 20 worker factors that can promote labour productivity of the labour-intensive works for road construction were subjected to PCA to assess their validity and reliability. The results report the suitability of the data to be analysed, the factor extraction and rotation, and the interpretation.
As shown in
Table 3, the KMO measure of sampling adequacy achieved a value of 0.892, exceeding the recommended minimum value of 0.7, and Bartlett’s test of sphericity was also statistically significant (<0.05), thus supporting the factorability of the data.
The pattern matrix in
Table 4 shows that, out of the initial 19 variables, PCA extracted 15 variables in four components with factor loadings above 0.4 with the potential to influence the labour productivity of labour-intensive works for road construction in Ghana.
As indicated in
Table 5, the eigenvalue was set at a conventional high value of 1.00 [
27]. In determining the number of principal components to be extracted, the latent root criterion was applied, which recommends that four components should be extracted because their eigenvalues are greater than one.
In
Table 6, a Cronbach alpha of 0.876 was obtained for the worker component. This satisfied the benchmark provided in [
26,
35], which states that the Cronbach alpha coefficient of a scale should be above 0.7. The closer the alpha (α) is to 1, the greater the internal consistency of items in the instrument is assumed to be.
Table 7 shows that after rotation, four components with eigenvalues exceeding 1.0 were extracted and were meaningful to retain. Factor one explains 36.43% of the total variance; factor two, 23.29%; factor three, 16.81%; and factor four, 6.79%. Thus, the final statistics of the PCA shows that three extracted factors explain a cumulative variance of approximately 83.32%.
Table 8 reveals the correlation of variables based on their factor loadings after rotation in PCA. Three components with eigenvalues above 1, as shown in
Table 4, were examined for the inherent relationships among the variables under each factor. Variables with the highest factor loading in one component belonged to that component; the highest factor loading had to be of a significant value of 0.4 or above (see
Table 4). Component 1 was labelled age of a beneficiary; Component 2 was labelled beneficiaries’ knowledge; Component 3 was labelled safety compliance; and Component 4 was labelled motivation of beneficiaries. The names given to these factors were derived from a close examination of the variables within each of the factors.
The scree plot presented in
Figure 1 also reveals the excluded factors by indicating the cut-off point at which the eigenvalues levelled off.
In
Table 7, four (4) components with eigenvalues greater than one were found using principal axis factoring. The following interpretations were produced in light of the evaluation of the underlying correlations between the variables under each factor. The age of beneficiaries, beneficiaries’ knowledge, safety compliance, and motivation of beneficiaries were the four categories under which Component 1 was classified. After carefully analysing the variables contained inside each of the elements, these component names were developed.
Below is an explanation of the individual indications that made up each of the three extracted components, along with a description of how each was characterized in detail during focus group sessions.
Table 7 shows the four (4) WC variables that were retrieved from the workers. The Component 1 measures were “employment of older beneficiaries from villages” (95.4%), “employment of young beneficiaries on projects” (91.4%), “incentives utilized to lure young people into sector” (77.6%), and “likelihood older beneficiaries will be replaced by younger people” (76.6%). The numbers in brackets represent the loadings on the corresponding factors. Approximately 36.4 percent of the overall variation may be attributed to this group.
The seven variables that were extracted from the worker component (WC) for Component 2 were as follows: “beneficiaries’ attitude towards the job they have to execute” (92.7%), “opportunities for employees to exercise their skills” (82%), “employees’ level of experience to do their work” (81.7%), “the number of multi-skilled project personnel in the company” (78.9%), “beneficiaries’ knowledge of scientific techniques” (71.7%), “Beneficiaries’ level of awareness of company policy” (70.1%) and “Beneficiaries’ having formal training in labour-intensive works” (56.5%). The numbers in parentheses denote the relevant factor loadings for that factor. This cluster was responsible for 23.3% of the total variation in the data. The results agree with the findings in [
41,
42,
43,
44,
45,
46,
47], which indicated that the knowledge of the workers on the task is very important when it comes to achieving productivity for road construction.
This group explained 16.8% of the total variation. The Component 3 variables that were retrieved from workers were as follows (
Table 7): “use of safety wear on-site” (95.2%), “level of safety reached on projects” (79.8%), and “degree to which safety requirements on a project meet with legislated criteria” (63.6%). The numbers in parentheses represent factor loadings.
The factors “quality of transportation facilities for workers” (85.8%), “management response to settle employees’ grievances” (85.4%), “the company’s incentive scheme for good performance” (78.7%), “promotion opportunities for employees” (71.71), “beneficiaries’ knowledge of career prospects” (64.5%), and “likelihood beneficiaries are paid on time” (44.7%) were the six (6) extracted worker component (WC) variables for Component 4, as shown in
Table 7. The numbers in parentheses represent the factor loadings. This cluster was responsible for 6.8% of the variation.
These results confirm the findings of the authors of [
48,
49], who stressed that there is a need for the workers to comply with construction specifications, ensuring the quality of the work delivered. Owing to the industry’s labour-intensive nature, the workforce factor plays a significant role in the construction project implementation process [
50]. The authors of [
51,
52,
53,
54,
55] confirmed that workforce accounts for 30–50% of the total project cost. Consequently, considering the worker component’s key role in achieving a higher level of productivity performance, construction professionals should pay more attention to the workforce dimension, which has four observable variables in line with the relevant findings: (i) age of beneficiaries [
56,
57,
58,
59]; (ii) beneficiaries’ knowledge [
60,
61]; (iii) safety compliance [
54]; and (iv) motivation of beneficiaries [
48].
This research is novel because it establishes a framework for measuring the productivity of workers in the labour-intensive sector of the Ghanaian construction industry by looking at variables such as the age of beneficiaries, the knowledge of beneficiaries, compliance with safety regulations, and the motivation of beneficiaries as illustrated in
Figure 2. A similar approach, based on the latent factors that were used to derive the labour productivity result variables, might be utilised to assess company-level productivity in the Ghanaian construction industry.
Construction industry experts might also utilise this information to better guide company-wide strategies. The study’s findings can also be used as a useful tool in planning to expedite the efficient utilisation of labour-intensive road building activity to boost productivity and meet contractual obligations. It can also help contractors estimate how long it will take to construct a certain road using the labour-intensive way.
This research has shown that the four components have a significant impact on the labour productivity of indigenous construction enterprises operating in Ghana. The findings could be used by the Association of Building and Civil Engineering Contractors of Ghana (ABCECG) to prioritise criteria for boosting the efficiency of local construction companies in Ghana. In the end, this will help the ABCECG determine where they need to focus their efforts in order to build up the capacity of their contractors.
As a reference, the framework can be used to make sure that road construction companies have all they need to have a highly productive workforce.