Safety Related Descriptive Statistics
Safety Related Descriptive Statistics
In research, both descriptive and inferential statistics are of paramount importance. Descriptive statistics help to summarize data into meaningful information, but does not allow the researcher to make further conclusions about the data. Inferential statistics helps make intuitive conclusions and even allows the researcher to test the hypothesis. This paper examines the tables provided with an aim of establishing whether or not there exists a relationship between the dependent and the independent variable.
In order to achieve this, the research paper is divided into three segments. The first segment defines the terms that are used throughout the paper. The second part examines the tables provided and draws conclusions. In the last part, the main points are restated as well as the hypothesis.
Keywords: descriptive statistics, inferential statistics, hypothesis, dependent variable, independent variable
Descriptive statistics refers to the analysis of data which helps the researcher to summarize data into a meaningful form. In short, descriptive statistics gives a description of the data but does not enable one to draw further conclusions from the data. In descriptive statistics, a combination of tables and graphical presentations, analyzes the data (Hirschberg, 2005).
On the other hand, inferential statistics makes use of samples which act as representatives of the whole population. In this case, various samples are picked using sampling methods that accurately represents the entire population. Inferential statistics helps the researcher to test the hypotheses (LAERD, 2012).
According to the Oxford Dictionary, a hypothesis is an explanation about a certain observation, a scientific problem or a theory which can then be further tested by statistical methods so as to either accept or reject it (Ashby, 2001).
In statistics, it is at times difficult to determine the relationship that exists between various variables. The researchers manipulate one of the variables so as to understand the effect of such manipulations on the other variable. The manipulated variable usually represents the independent variable whereas the dependent variable is determined depending on what effect the manipulation of the independent variable has. Some researchers refer to dependent variables as explained or determined variables while the independent variables as regressors (Ashby, 2001).
The research in question was undertaken to determine whether there was a direct relationship between the injury rate in work environments and the safety climate. The researcher wanted to determine the effect of varying the safety climate on the injury rate. However, research indicated no relationship between the two variables. This was clearly portrayed by the line graph provided. It showed neither positive nor negative correlation between the dependent and independent variables.
After a clear examination, it was noted that there was no relationship between the safety climate and the injury rate. From the line graph, shown, it would be right to accept the hypothesis that indeed, the two variables are not correlated. The manipulation of the independent variable had no effect on the dependent variable and therefore, the null hypothesis was accepted. In short, from the analysis conducted, the alternative hypothesis was rejected.
In line with this, a regression analysis was carried out which indicated similar results. The analysis of variance table generated a p value of 0.930 greater than the tabular value which was not at all significant. The alternative hypothesis was thereby rejected, and the null hypothesis accepted (TREK, 2012).
Background Information and Objective
In any organization, it is particularly important to ensure the safety of the workers. Most organizations, particularly in Canada have significantly been able to reduce the injury rates in work places. However, the workers are still injured it would therefore be important to continue devising ways to avoid injuries in work places. One of such method of reducing injuries is to ensure a tighter safety climate. Safety climate is defined as merged efforts by workers and organization in ensuring low levels of injuries within the organization. Research indicates that, safety climate has a remarkably passive effect of the health standards of the workers since, it ensures that injury levels and even possible deaths are reduced to a great extent. Studies indicated that if companies progressively monitored their safety climate, then the injuries would be significantly reduced in such organizations (IWH, 2007).
Objective
This research was conducted in three locations: Boston, Phoenix and Seattle. The research aimed at determining the effect of enhancing safety climate on the injury rates of the workers. The researcher expected that there would be a visible relationship between the two variables.
Hypotheses
As outlined earlier on, the aim of this paper was to determine the relationship between the safety climate of the employees in the three locations and their injury rates. Before the researcher could proceed with data collection, it was indispensable to outline the hypothesis that could later be tested. Such a hypothesis would either be a positive statement that would go in line with the various theories about safety climate and the injury rates or a negative statement that would go against the theories. The positive hypothesis is called an alternative whereas a negative statement is called a null hypothesis. After the researcher has collected samples which are representative to the population, he proceeds with further analysis so as to make a generalization about the entire population (Lehmann, 2007).
In research, the null hypothesis is denoted by H0 while an alternative hypothesis is denoted by HA. In this case therefore, the hypotheses were as follows:
Alternative Hypothesis (HA): There is a relationship between the safety climate and the injury rates.
Null Hypothesis (H0) : There is no relationship between the safety climate and the injury rates.
As clearly indicated above, the researcher aimed at rejecting the null hypothesis, thereby accepting the alternative hypothesis or accepting the null hypothesis hence rejecting the alternative hypothesis. It is at times difficult to draw conclusions from descriptive statistics and therefore, had to use the inferential statistics that would assist in making generalizations of the population from the various samples picked by the researcher.
Descriptive Statistics (Table 1)
Number Range mean Standard Deviation Variance
Statistic Statistic Statistic Std Error Statistic Statistic
Num Emps 51 40 24.02 1.05 7.495 56.18
Hours Worked 51 83200 49960 2183.07 15590.23 2.431
Per Safe Beh 51 .58 .8658 .01946 .13895 .019
SafetyClimate 51 4.3 4.697 .1449 1.0350 1.071
Risk 51 6 4.59 .282 2.012 4.047
ExperiencedCoded 51 2.00 1.9608 .11534 .82367 .678
Valid N) (listwise 51
Analysis of Variance Table (2)
Model Sum of squares Degree of Freedom Mean Square F Sig.
Regression 2.453 1 2.453 .008 .930a
Residual 15265.765 49 311.546
Total 15268.217 50
Bar Graph Generated From the Plotted Line Graph
Presentation, Interpretation and Discussion of Results
From the above bar graph generated from the plotted line graph, it was deduced that as one increases the safety climate, there was an effect on the injury levels. However, there was no consistency since, with a safety climate level of two, the injury level was zero. As the safety level was increased to three, the injury level raised up to twenty. Further, the safety level was increased to seven , and the injury level was reduced close to zero. According to the background information provided, it would be expected that the two variables are inversely proportional to one another. This means that, with an increase in safety level, the injury rates in the organization should significantly decrease. This is in line with the alternative hypothesis that shows a relationship between the safety level and the injury rates. However, results indicate an wholly different scenario that conforms to the null hypothesis earlier on stated in the paper.
The research further, analyzed the data but in this case, using a sample size of fifty with a degree of freedom= 1. The injury rates were regressed against the safety climate. However, the line of best fit could not be established owing to the fact that there was no relationship between the two variables. In short, if the safety climate values are manipulated, there is no consistent effect on the determined variable. In line with this, an analysis of variance table was generated which further attested to this. In order to accept the alternative hypothesis, the calculated Fishers test should be greater than the tabular value and more particularly more than twice the tabular value. Since the tabular value was not provided, the p value was used which was outside the five per cent level of significance or the ninety five percent confidence interval. The P value was greater than 0.05 and therefore, not at all significant.
The line graph provided did not show any visible relationship between the two variables. This means that even if the safety climate were to be further increased, the researcher could not predict the effect on the explained variable.
Conclusions
From the above deliberations, it is exceedingly clear that there exists no visible relationship between the safety climate and the injury rate. Theoretically, it would be expected that the relationship exists. The data collected from the three locations did not conform to the theoretical framework. However, the writer of this paper had a strong feeling that indeed, there is a relationship between the two variables. The researcher ruled out the effects of other factors that could have a significant effect on the injury levels apart from the safety climate.
The other factors that affect injury rates in an organization should not be ruled out but incorporated in the model. Possibly, if the other factors were incorporated, then the relationship between the two variables could be determined.
References
Ashby. (2001). Students Dictionary. New York: Oxford University Press.
IWH. (2007). Safety Climate shows Promise in Injury Prevention. At Work .1-4. Retrieved from. www.sciencedirect.com/sciencearticle/pii.
LAERD. (2012). Descriptive and Inferential statistics. Statistics . 1-2. Retrieved from. https://statistics.laerd.com/statistical
/descriptive-inferential-statistics.
Lehmann. (2007). Testing Hypothesis. Statistics . 3-8. Retrived from. cran.r-project.org/doc/contrib/Krijnen-IntroBioInfStatistics.pdf.
TREK, S. (2012). What is Hypothesis Testing. 1-5. Retrieved from. stattrek.com/probability-distributions/binomial.aspx.
Hirschberg, J., Lu, L., & Lye, J. (2005). Descriptive methods for cross-section data. Australian Economic Review. 38 (3). 333350.
Trochim, W. (2006). Web center for social research methods: Selecting statistics. Retrieved from http://www.socialresearchmethods.net/
