Sociodemographic characteristics
Of 401 people surveyed, 397 (99%) responded. The majority of the respondents were between 25-30 years of age (232, 58%) and female (290, 73%). Fifty-one percent of respondents had greater than 4 years of work experience, and 52% of the respondents were HEWs (Table 2). Figure 1 plots the responses to each motivational task, where 1 represents “strongly agree” and 5 represents “strongly disagree”, and blue dots represent items where a lower score is a priori better (e.g. “I am respected in my community for the work I do”) and red dots where a lower score is worse (e.g. “I intend to stop working in this role in the next 12 months”).
Motivation level
Of the total respondents, 61% were motivated, responding “very good” or “excellent” (95% CI 57% - 66%). The most motivating factor mentioned by 70% of the participants was the opportunity to improve health. The most demotivating factor mentioned by 29% participants was workload.
Motivation construction – Exploratory factor analysis
We ran a factor analyses between one and five factors, and ultimately found that the three-factor model fit the data best. We removed 11 items which did not load to 0.4 on any factor. We calculated the final factor score by multiplying the items by the factor score and summing. Table 3 shows the factor loadings. In addition to the factor loading above 0.4, we used eigenvalues >1 as a criterion to reduce the number of factors into three. We labelled the three factors as personal and altruistic goals; pride and personal satisfaction; and recognition and support. The minimum and maximum score for personal and altruistic goals was 1 and 4.85; for pride and personal satisfaction it was 1 and 4.17; and for recognition and support, it was 1 and 4.5.
Factors related to overall motivation
Table 4 shows the results of the regression model with overall motivation as the outcome variable and a number of explanatory factors fitted.
The overall motivation score mean was 2.18 out of 5 (95% CI 2.1, 2.27; P=0.001). Job satisfaction was associated with overall motivation, where an increase of 1 in job satisfaction was associated with an increase in motivation of 0.23 Likert points (95% CI 0.13, 0.34; P=0.001). We found variation in motivation across regions, where participants from SNNPR reported lower motivation than the Amhara region by 0.35 Likert points (95% CI 0.12, 0.59; P=0.003). Participants who have a medium workload (meaning that they have enough time to complete duties) were also less motivated than participants who have a light workload by 0.48 Likert points (95% CI -0.90, -0.06; P=0.024), though high workload was not significantly different. There was no significant variation in overall motivation job title and perceived gross salary.
Factor 1: Personal and altruistic goals
The first factor identified items which related to personal success and goal setting and altruism. We refer to this factor as “personal and altruistic goals”. The mean response for personal and altruistic goals was 1.6 Likert points (95% CI 1.6, 1.7). From the results of the regression analysis summarized in Table 5, an increase in 1 in job satisfaction was associated with an increase in personal and altruistic goals score of 0.34 Likert points (95% CI 0.29, 0.39; P<0.001). Variation in personal and altruistic goals was seen among regions, where participants from Oromia and SNNPR had lower personal and altruistic goal scores than Amhara region by 0.13 Likert points (95% CI -0.25, -0.02; P=0.018) and 0.12 Likert points (95% CI -0.23, -0.01; P=0.039) respectively. Annual leave (time out of work) was another significant factor associated with personal and altruistic goals, where an increase of 1 was associated with a decrease in personal and altruistic goals of 0.08 Likert points (95% CI -0.12, -0.04; P=0.001). Age, gender and perceived salary fair had no significant association with personal and altruistic goals.
Factor 2: Pride and personal satisfaction
The second factor consisted of variables relating to the pride and personal satisfaction respondents experienced in their jobs, and the satisfaction their jobs gave them. The mean score of pride and personal satisfaction was 1.78 Likert points (95% CI 1.72, 1.84). A regression analysis in table 6 indicated that pride and personal satisfaction score vary among regions with SNNPR being significantly low by 0.11 Likert points (95% CI -0.22, 0.001; P=0.051) as compared to Amhara region. Likewise, as leave (time out of work) increased by 1, pride and personal satisfaction score decreased by 0.06 Likert points (95% CI -0.1, -0.02; P=0.003). Age was significantly associated with pride and personal satisfaction. There were significantly high scores among those older than 30 years old by 0.14 Likert points (95% CI 0.01, 0.28; P=0.049) as compared to those between the ages of18-24 years old. Another factor affecting pride and personal satisfaction was job satisfaction. As job satisfaction increased by 1, pride and personal satisfaction score increased by 0.55 Likert points (95% CI 0.50, 0.60; P=0.001). Gender and work experience did not have significant association with pride and personal satisfaction.
Factor 3: Recognition and Support
Factor 3 synthesised survey items focusing on the recognition and support that participants received from colleagues and seniors. The mean score of recognition and support was 1.8 (95% CI 1.7, 1.8; P=0.001). From the regression analysis indicated in Table 7, factors associated with recognition and support were job title, region, leave (time out of work) and job satisfaction. Recognition and support scores were significantly high among health care providers as compared to HEWs by 0.12 Likert points (95% CI 0.001, 0.25; P=0.053). Compared to Amhara region, recognition and support scores were significantly low among SNNPR by 0.15 Likert points (95% CI -0.32, -0.03; P=0.017) and Oromia region by 0.17 Likert points (95% CI -0.29, -0.01; P=0.034). Another factor affecting recognition and support scores was job satisfaction. As job satisfaction increased by 1, recognition and support score increased by 0.35 Likert points (95% CI 0.29, 0.41; P=0.001). However, as average leave days increased by one, recognition and support scores significantly decreased by 0.09 Likert points (95% CI -0.14, -0.04; P=0.001). Age, work experience, gender and perception of fair salary has no significant association with recognition and support.