Factors Affecting Employee Weekly Income: A Study On Qualification, Age, Working Hours (Essay).

Learning outcomes

1.Use appropriate methods of sampling and describe suitable methods of collecting data.

2.Generate appropriate graphs and descriptive statistics and use them to describe features of various datasets.

3.Find probabilities from various probability distributions.

4.Explore sampling distributions and use appropriate and valid tests of significance and confidence intervals for means and proportions.

5.Undertake linear regression analysis, and interpret the results.

6.Clearly communicate the results of data analysis.

Primary and secondary sources of data collected

Income is the revenue generated from selling goods and services (Haraldsson, 2017).  (DAVID, 2011) Define it as the money that an individual receives in compensation for his or her labor and services. (Jones, 2010) Studies on the factors that affect income reveal that qualification is a measure of one’s potentiality and therefore, highly skilled personnel tend to generate more income in both the private and public sector. Another study conducted in New Zealand reveals that qualification is directly proportional to income (Sheree, 2012) whereby the greater the level of qualification the higher the income. The time that individuals spent in doing work have a great influence on the weekly income as revealed by (Asafa, 2015). However, studies by (Volkoff, 2010) indicate that older people tend to spend less time in work because of their vulnerability based on depression, tiresome, and memory capacity to handle multiple-task at a time and with speed, therefore, presenting less income compared to young people. (Adjei, 2018) in his research studies where he targeted respondents aged between 50 to 75 years revealed that the proportion of men working at the age of 50 to 59 years is higher than the proportion of women, but this trend tends to change as they grow older, however, their output based on income tends to decrease as their working time tend to decrease. There are several factors affecting the income as is dependent on the qualification of the laborers, age and the time a person spent to work. Lack of sufficient data (Fox, 2012) is the greatest challenge facing both private and public sectors in measuring the aforementioned factors. Availability of data is significant in effectively estimating the factors that affect employee income (Kasra, 2012).

Since income is generated from sales or as the compensation for the labor and services, it is the responsibility of an individual in either private or public firm to maximize the income by putting into consideration factors like employee qualification, age, and the time in hours spent in work. The fact that little research has been conducted in companies on factors affecting companies and employee income, it is worth carrying out the research.

Every business, as well as individuals, have a motive or an aim to maximize income and minimize expected losses (Yamarone, 2017), whereas this is true, to achieve this goal; several businesses have put in place strategies believed to help achieve their goals.  This research was therefore conducted in order to investigate the factors that affect the income in order to come up with suitable recommendations for improving the weekly income.

1. What factors affect the weekly income of an employee?
2. How do these factors being investigated affect the weekly income of an employee?
3. To increase or improve existing income, what must or should it be done?

Selection of the best model using LASSO and LAR techniques

The general objective of this study is to determine the factors that affect employee weekly income.

1. To investigate if there is a linear relationship between the time spent on work and the weekly income.
2. To identify the effect of the qualification on employee weekly income.
3. To identify the effect of age on employee weekly income.
4. To investigate how gender and ethnicity affect employee weekly income

A survey refers to the process of data collection on various aspects or characteristics of a study population. This research will make use of an online survey data that was presented in form of questionnaires and administered online to the target population. The respondents received the questionnaires via their emails, filled and returned them.

An online survey is the most effective method of data collection for this research as it allows for freedom of filling the questions at respondents’ own pleasure and also it is less cost effective.

2.2. Data description

Both primary and secondary sources of data collected were used in this research. An online survey was employed for collecting primary data while company websites https://new.censusatstudent.org.nz/resource/nz-incomes-surf/ was used for collecting secondary data.  The sample size is 100 individuals consisting of both male and female that was obtained through random sampling. In addition, the targeted respondents are aged between 25 to 64 years. The data variables are age, sex, ethnicity, qualification, and income.

LASSO is a shrinkage and variable selection method for linear models. It is a regression analysis technique that performs both variable determination and regularization so as to improve the forecast exactness and interpretability of the factual model it produces. It is an extension of linear regression using shrinkage.

A Stepwise method was used for selecting the best model. This is a method of fitting regression models in which the choice of predictive variables is carried out by an automatic procedure. In each step, a variable is considered for addition to or subtraction from the set of explanatory variables based on some predetermined criterion.

The method shall select variables to be included in the model one after another.

Model 1: income = age

Model 2: income = age + sex

Model 3: income = age + sex + qualification

Model 4: income = age + sex + qualification + time in hours

Model 4 is the best lasso model as it includes only the variables that are statistically significant.

LAR is a technique of fitting a regression model for a linear combination of a subset of potential covariates. The calculation is like forward stepwise regression, however as opposed to including factors at each progression, the evaluated parameters are increased toward a path equiangular to everyone's relationships with the residual.

To select the best method, we use forward regression. This is a technique which involves starting with no variables in the model, testing the addition of each variable using a chosen model fit criterion. We found that the model below is fit for this study.

Population demographic analysis - age, sex, and ethnicity

Income = age + sex + qualification+ time in hours

Given the models selected from lasso and LAR, we shall investigate the following hypotheses;

1. H₀: Age does not affect employee weekly income.

H₁: Age does affect employee weekly income.

1. H₀: Sex does not affect employee weekly income

H₁: Sex does affect employee weekly income

• H₀: Qualificationdoes not affect employee weekly income

H₁: Qualification does affect employee weekly income

1. H₀: Time in hoursdoes not affect weekly employee income.

H₁: Time in hours does not affect weekly employee income.

Based on the two information criterion lasso and LAR, we shall investigate the factors that affect the income. The following variables shall be examined on their effect on income; age, sex, qualification, and time in hours that an employee or individuals spent working in a week.

The population was aged between 25 to 60 years of age. Most population 23 representing 23% were aged between 30-34 years followed by 19 representing 19% who were aged between 25 to 29 years. The lowest number of respondents 7(7%) of respondents from the data were aged between 60 to 64 years.

Age Frequency Table
		Frequency	Percent	Valid percent	Cumulative percent
Valid	25-29	19	19.0	19.0	19.0
	30-34	23	23.0	23.0	42.0
	35-39	12	12.0	12.0	54.0
	40-44	8	8.0	8.0	62.0
	45-49	13	13.0	13.0	75.0
	50-54	9	9.0	9.0	84.0
	55-59	9	9.0	9.0	93.0
	60-64	7	7.0	7.0	100.0
	Total	100	100.0	100.0

A total of 100 individuals were sampled to take part in the study. Out of the 100, 47 individuals making 47% were male while 53 representing 53% were female.

Sex Frequency table
		Frequency	Percent	Valid percent	Cumulative percent
Valid	Male	47	47.0	47.0	47.0
	Female	53	53.0	53.0	100.0
	Total	100	100.0	100.0

Majorities of the population 23(23%) were aged between 30 to 34 years of age with a high percentage of the male being 13(56.52%) and female10 (43.48%). These, therefore, reveal that the majority of people working to obtain their source of income lie between the ages of 30-34 years. The number of men decreases as they grow older as observed in the age between 60-64 years where male and female were 3(42.86%) and 4(57.14%) respectively. This supports a study that was conducted by (Corneel, 2010) which revealed that women tend to become more active than men as they grow older and therefore, women tend to work for many years than men.

Cross-tabulation of age and sex

		Sex
		Male	Female
		Count	Count
Age	25-29	9(47.37%)	10(52.63%)
	30-34	13(56.52%)	10(43.48%)
	35-39	8(66.67%)	4(33.33%)
	40-44	2(25%)	6(75%)
	45-49	4(30.77%)	969.23%)
	50-54	3(33.33%)	6(66.67%)
	55-59	5(55.56%)	4(44.44%)
	60-64	3(42.86%)	4(57.14%)

The data from the population which was sampled from 100 individuals shows that most of the respondents 78 representing 78% were from Europe followed by 11 representing 11% were from Mauritania.

Ethnicity
		Frequency	Percent	Valid percent	Cumulative percent
Valid	Europe	78	78.0	78.0	78.0
	Mauritania	11	11.0	11.0	89.0
	Other	7	7.0	7.0	96.0
	Non-Mauritania	1	1.0	1.0	97.0
	Mauritania combination	1	1.0	1.0	98.0
	Pacific	2	2.0	2.0	100.0
	Total	100	100.0	100.0

Correlation analysis is carried out to assess the relationship between variables. Of more interest in this analysis, is the correlation between the dependent and independent variables

Results depict that ethnicity does not show any significant association with the dependent variable, income since its p-values exceed the critical value, 0.05. The variables age, sex, qualification and hours spent in work depict a statistically significant relationship between them and the dependent variable, income.

The Pearson correlation coefficient between sex and income is -0.391 implying a negative association between the two variables. A change in sex in one direction results to a change in the income in the opposite direction.

Correlation analysis of variables

The Pearson correlation coefficient between age and income was found to be 0.122 implying that there exists a positive association between age and income. This positive association can be interpreted as; an increase in age by one-year results to a corresponding increase in income by about 0.122 units. This can be attributed to experience gained in particular work as time goes by.

The Pearson correlation coefficient between qualification and income was found to be 0.198 implying a positive association between employee qualification and income generated. The 0.198 association coefficient can be interpreted as; an increase in the number of qualified employees by one result to a corresponding increase in income by about 0.198 units.

The Pearson correlation coefficient between time in hours and weekly income was found to be 0.405 implying a positive association between time and income generated. The 0.405 association coefficient can be interpreted as; an increase of time by one-unit results to corresponding increases in income by about 0.405 units.

Correlations
		Income	Age	Sex	Ethnicity	Qualification	Hours
Pearson correlation	Income	1.000	.122	-.391	-.074	.198	.405
	Age	.122	1.000	.092	.132	-.216	.081
	Sex	-.391	.092	1.000	-.107	-.177	-.407
	Ethnicity	-.074	.132	-.107	1.000	-.090	.101
	Qualification	.198	-.216	-.177	-.090	1.000	-.088
	Hours	.405	.081	-.407	.101	-.088	1.000
Sig. (1-tailed)	Income	.	.113	.000	.231	.024	.000
	Age	.113	.	.181	.096	.015	.212
	Sex	.000	.181	.	.144	.039	.000
	Ethnicity	.231	.096	.144	.	.186	.159
	Qualification	.024	.015	.039	.186	.	.191
	Hours	.000	.212	.000	.159	.191	.
N	Income	100	100	100	100	100	100
	Age	100	100	100	100	100	100
	Sex	100	100	100	100	100	100
	Ethnicity	100	100	100	100	100	100
	Qualification	100	100	100	100	100	100
	Hours	100	100	100	100	100	100
**. Correlation is significant at the 0.01 level (1-tailed).
*. Correlation is significant at the 0.05 level (1-tailed).

We examined the following model and its variables.

Income = Age + Sex + Qualification+ Time in hours

Except for ethnicity, the independent variables age, sex, qualification, and time in hours were found to be statistically significantly associated with income. This is because the p-values for the association between the variables were less than 0.05 implying a rejection of the null hypothesis that there is no association between the dependent and independent variables. Age, qualification, and time in hours depict a positive association with the dependent variable while sex was negatively associated with income

Correlations
		Income	Age	Sex	Qualification	Hours
Pearson correlation	Income	1.000	.122	-.391	.198	.405
	Age	.122	1.000	.092	-.216	.081
	Sex	-.391	.092	1.000	-.177	-.407
	Qualification	.198	-.216	-.177	1.000	-.088
	Hours	.405	.081	-.407	-.088	1.000
Sig. (1-tailed)	Income	.	.113	.000	.024	.000
	Age	.113	.	.181	.015	.212
	Sex	.000	.181	.	.039	.000
	Qualification	.024	.015	.039	.	.191
	Hours	.000	.212	.000	.191	.
N	Income	100	100	100	100	100
	Age	100	100	100	100	100
	Sex	100	100	100	100	100
	Qualification	100	100	100	100	100
	Hours	100	100	100	100	100
*. Correlation is significant at the 0.05 level (1-tailed).
**. Correlation is significant at the 0.01 level (1-tailed).

Regression analysis was conducted including all the variables involved in model 4 LAR and LASSO.

Model summary
Model	R	R square	Adjusted r square	Std. Error of the estimate
1	.550a	.303	.266	403.65438
A. Predictors: (constant), hours, age, ethnicity, qualification, sex

R square equals 0.303 implying that the independent variables explain about 30.3% of the dependent variable.

ANOVAa
Model		Sum of squares	Df	Mean square	F	Sig.
1	Regression	6659306.192	5	1331861.238	8.174	.000b
	Residual	15316064.808	94	162936.860
	Total	21975371.000	99
A. Dependent variable: income
B. Predictors: (constant), hours, age, ethnicity, qualification, sex

ANOVA regression test had a p-value of 0.000 implying that it would be statistically significant to include an ANOVA model.

Coefficientsa
Model		Unstandardized coefficients		Standardized coefficients	T	Sig.
		B	Std. Error	Beta
1	(constant)	415.244	276.672		1.501	.137
	Age	38.084	18.584	.183	2.049	.043
	Sex	-241.105	92.017	-.257	-2.620	.010
	Ethnicity	-66.542	41.964	-.139	-1.586	.116
	Qualification	80.130	34.956	.208	2.292	.024
	Hours	10.172	3.084	.318	3.298	.001
A. Dependent variable: income

The independent variable to be included in the prediction model include age, sex, qualification, and time(hours) since their p-values are less than 0.05 implying that they would be statistically significant in the model.

Results of the analysis have shown that several factors affect weekly income. The factors are as discussed below;

1. Age

Results of the regression analysis show that age explains about 38.084 times of the dependent variable (income). The more an individual grows older, the more experience one gain and subsequently more income generated (Burns, 2015).

However, in his study on the active population between young and old adult, a conflicting result was obtained by (Zell, 2014). He failed to find any statistical association age and income.

1. Sex

Significance of qualification, age, and working hours in affecting employee income

Results of the regression analysis depict the negative effect of sex on plasma beta-carotene with a coefficient of - 241.105. This is an implication that a change in gender would result in a change in income in the opposite direction by about 241.105 units. Since the female gender is represented by a higher number compared to the male gender, it would be right to argue that the female gender shall have a negative impact on income. Similar findings have been found by previous researchers, including (Lo Sasso, 2011) and (Sueuk, 2016) they both argued that women were negatively associated with income.

Contradicting results were obtained by (Stoet, 2013) found out that there exists no association between sex and income. These results were almost similar to those observed by (Vandelanotte, 2010). Vandelanotte observed an association coefficient of -0.003 between gender and income.

• Qualification

The findings have depicted high qualification has positive impacts on income with a regression coefficient of 80.130. This implies that an increase employee with high qualification leads to an increase in income. Qualification explains about 80.13 times of the dependent variable. (Olckers, 2015) Studies reveal that the highly skilled personnel tend to produce a better result and subsequently better income; which agrees with findings of the current study.

1. Time(hours)

The study found out a positive relationship between time in hours and weekly income with a regression coefficient in 10.172. This implies that an increase in time that employees work by 1 hour would lead to an increase in weekly income by 10.172 units. Consequentially, a decrease in time that employees work by 1 hour would result in 10.172 units decrease in weekly income. Our current findings are similar to the findings of (Kolodziejczyk, 2012)

Contradicting findings have previously been shown by (Jarousse, 2010) who failed to find any significant relationship between the two variables. He found a significant value of 0.63 which is greater than the critical point 0.05.  He concluded that working less time and increase earning.

Conclusion

Factor selection is crucial and challenging in this field of study, mainly because the desired output varies for a different set of data, and it is hard to find a model that works for every kind of problem. For these reasons, the present study made use of the LAR model selection techniques. The techniques helped us to identify the model with the most relevant features (variables) for the dataset to analyze. The task becomes even more challenging when dealing with high-dimensional datasets.

No significant association found between ethnicity and income

The findings of the present study have shown that a number of factors affect weekly income. Among these factors are age, sex, employee qualification, and time in hours that an individual spent in working. These factors have a different impact on the dependent variable (weekly income). High qualification, time in hours and age were found to positively affect employee weekly income whereas sex had a negative impact on employee weekly income. Proper control measures should, therefore, be taken to control for these factors in order for companies and individuals to increase their weekly income.

Apart from the independent variables being investigated in this research, there are other factors which involve managerial and political factors which affect the dependent variable income and was not discussed in this research. Furthermore, there is little research that has been conducted relating factors affecting employee income.

From the results of the analysis of this research, the following are the recommendation which should be employed to increase the income by the employees or companies both public and private

1. Employers should put qualification into consideration during employment as this factor correlate positively with income.
2. Ethnicity is not a factor to consider during employment at all. This variable is not statistically significant rendering it unnecessary.

• Employees and individuals from both the public and private sector should increase their time to work to increase their weekly income. Time spent in a week to work is positively correlated to weekly income,

1. Age is a factor which is positively correlated with income. This is due to experience gained every year by the employee. Experienced employees should be given priorities for employment than inexperienced people in order to increase weekly income

References

Adjei, N. K. (2018). The Capability of multi-tasking among the age population. International Journal for Equity in Health, 2-12.

Asafa Jalata, H. F. (2015). Current Perspectives in Social Theory. Capitalism and the Common Sense of Time and Money, 29-74.

Burns, W. (2015). Textile-Led Design for the Active Ageing Population. Experiences in the design, iterative development, and evaluation of a technology-enabled garment for active ageing

Vandelanotte, M. J. (2010). Australian and New Zealand Journal of Public Health. Physical activity trends in Queensland (2002 to 2008): are women becoming more active than men?, 248-254.

Current Perspectives in Social Theory. (2015). Capitalism and the Common Sense of Time and Money, 29-74.

DAVID L. SJOQUIST, A. V. (2011). Public Budgeting and Finance. The Impact of Tax Revenue from Capital Gains Realizations on State Income Tax Revenue and Budget Conditions, 31-48.

Fox, C. K. (2012). A Lack of Strategies for Effective Decision Making. insufficient data, 33-38.

Haraldsson, M. (2017). When revenues are not revenues: the influence of municipal governance on revenue recognition within Swedish municipal waste management. Local Government Studies, 669-688.

Jarousse, J.-P. (2010). Economics of Education Review. Working less to earn more: An application to the analysis of rigidity in educational choices, 50-62.

Jones, J. P. (2010). Factors Affecting Income. The Journal of Higher Education, 185-190.

Lo Sasso, A. T.-F. (2011). The Unexplained Trend Of Men Earning More Than Women. Pay Gap For Newly Trained Physicians, 194-212.

Olckers, C. P. (2015). European J of International Management. Psychological ownership as a requisite for talent retention: the voice of highly skilled employees, 53-59.

Volkoff, C. B. (2010). Applied economics. Does intense time pressure at work make older employees more vulnerable? A statistical analysis based on a French survey, 754-762.

Sheree J. Gibb, D. M. (2012). Childhood family income and life outcomes in adulthood: Findings from a 30-year longitudinal study in New Zealand. Social Science & Medicine, 8.

Stoet, G. O. (2013). Are women better than men at multi-tasking? BMC Psychology, 2-10.

Sueuk, U. P. (2016). Men earn substantially more than women. Board & Administrator for Administrators Only, 5-9.

Yamarone, R. (2017). Maximize Profits and Minimize Losses. Journal of economic, 23-28.

Zell, E. B. (2014). Social Psychological and Personality Science. ou May Think You're Right, Young Adults Are More Liberal Than They Realize, 327-338.