Descriptive Statistics and Central Tendency - Article Example

Descriptive statistics and Central tendency Descriptive statistics and Central tendency Introduction Descriptive statistics are techniques that are used to summarize and present the most important aspects in large quantitative information. Such summaries may be in form of graphs, tables, or charts. On the other hand, central tendency is a term used to refer to the number among numeral information that best recapitulate the entire information and centrally placed (Macfarland, 1998). The central tendency of quantitative information is obtained by finding its mean, mode, median, variance, standard deviation, or its rage. The mode is the most repeated score in the collected data. The median is the mid number depending on the arrangement of the collected data. The mean is the arithmetic average while variance, standard deviation, and rage give the dispersion, variation or spread of other variables from the average score. The study The paper will focus on the article “The Effect of Evidence-Based Practice on Workplace Empowerment of Rural Registered Nurses” by Bleden, Leafman, Nehrenz, and Miller. The main purpose of the study was to find out the levels of evidence based practice and intellectual capacity of the rural nurses as well as, their levels of empowerment at the workplace. The study was carried out though an online survey where only 42 rural nurses out off the 500 randomly selected nurses, in a population of 25,000 nurses, completed the survey (Belden et al, 2012). The nurses were expected to complete three questioners. The first consisted of age, gender, educational background, and employment status. The second one aimed at evaluation the nurses’ knowledge, attitudes, and levels of self-esteem founded on evidence based practice. Lastly, the third section consisted of a Psychological Empowerment Instrument of 12 self scored items that evaluates ones perception towards employee empowerment. The result portrayed a close relationship between the evidence based practice used and levels of empowerment at workplace. Regression analysis further denoted educational background as the main variable that affects the evidence based practice used and empowerment levels (Belden et al, 2012). The ability of the rural nurses to offer quality and reliable healthcare services can be enhanced by ensuring that evidence based practices are in concurrent with empowering work structures implemented. Central tendency used in study One of the central tendency measures was calculating the mean of the evidence base practices from the data collected, as well as, in other categories (Saline, 1999). The other technique used in each case was finding the standard deviation from the mean. For instance, the values achieved using the 7-item Likert scale evidence base practices categories, had its means tabulated in the table below. In this case, coming up with the mean as the central tendency measures is inappropriate and does not make sense. The fact that the results were based on a 7-item Likert scale, makes them subjects’ opinion and inaccurate. Results obtained in from subjects’ opinions are referred to as ordinal values and not real numbers. For instance, rating the EBPE attitude 5 does not show the same sense of agreement with rating it as 6 or 7, neither is adding two disagrees (3) amount to one agree (6). Additionally, the same person may have marked 5 in two different questions but with different feeling on each question (Saline, 1999). This is a mistake commonly observed in computed data where data is directly fed into a computer. This is because computer automatically gives the means of all fed values, which is the case in this study. However, this can be avoided by knowing the kind of data being analyzed, a property that a computer program cannot automatically do. That said, the best data to calculate its mean is a continuous data of intervals and ratios. Assumptions of the central tendency measure Assumptions in central tendency measure are illustrations of when and which central tendency measure to use. They entirely dependent on the kind of variables collected during a study. These kinds include nominal, ordinal, interval, or ratio variables. In the case of nominal variables, only the mode can be used and neither the mean nor the median. This is because the numbers cannot be organized in any manner to find the median while the mean is impossible to find because the numbers are only representatives and not real number (Macfarland, 1998). In ordinal variables, it is inappropriate to calculate the mean but appropriate for the mode and the median. Finally, the interval and ratio variables are open to all the three central tendency measures. In this study, not all the central tendency measure assumptions were met. For instance, it was inappropriate to calculate the mean of variables achieved using the 7-item Likert scale. In such a case, the best central tendency measure could have been finding the mode and the median of the variables and not the mean (Saline, 1999). The study on evidence based practices and empowerment level of the nurses at their working place is a qualitative research that cannot be measured in real numbers, hence it is senseless to calculate the mean. Levels of measurement described and evaluated Levels of measurement are aspects that show the relationship of values assigned to a certain attribute to bring out the variation between them. There are four levels of measurement which are; nominal, ordinal, interval, and ratio. Nominal level measures values that indicate only the attributes of an object in a unique way (Trochim, 2006). Ordinal level measures the attributes as well, but measurable attributes such educational levels. In Interval and ratio levels, the variation between the attributes is considered to make sense and can be used in analysis and evaluation. In this study, the nominal and ordinal levels of measurement are incorporated. The nurses are asked to rate their gender, knowledge, attitude, and skills. All these attributes cannot be measured using numbers. Therefore, this aspect shows that the study can only be classified at nominal level (Trochim, 2006). Additionally, there are attributes in the study, such as, education level and age that qualify as ordinal variables. Therefore, it is clear that since both ordinal and nominal variables are not real, one cannot calculate their mean. Data displays, strengths and weakness Data collected from the 42 completed surveys in the study was tabulated in an excel Microsoft formulated table. Each of the variables was entered against the attributes in the study as categorized in the questioners. The data was analyzed using descriptive statistics techniques and central tendency measures. The results from the study were displayed in graphs, tables, and pie charts. This form of descriptive display of data observed from the study has both advantages and disadvantages (Madrigal and Mcclain, 2012). In this study, the tables and graphs provide direct information about the variables. For instance, the table directly indicates whether one is a male or female, his or her actual age group. Secondly, tables simplify the information into more understandable and flexile form. Lastly, tables create opportunity to identify unanticipated situation in the future. On the other hand, the only weakness portrayed in the tables and graphs was that it only represented 8 percent of the intended participant in the survey. In conclusion, it is clear that the assumptions of statistics and central tendency measures in evaluating evidence based practice and empowerment levels at work place were not met in this case. The type of data and level of measurement used in this study was not in conjunction with the descriptive statistical techniques and central tendency measures. The analyst was not keen to recognize the difference between qualitative and quantities variables. Finally, the analyst used the inappropriate central tendency measure, instead of finding the mode and the median of the data, the analyst calculated the mean of the data. This was inappropriate because it impossible to calculate the mean of nominal and ordinal variables. References Belden V. Catherine, Leafman Joan, Nehrenz Guy, and Miller Patricia. (2012). The Effect of Evidence-Based Practice on Workplace Empowerment of Rural Registered Nurses. Journal of Rural Nursing and Health Care, 12(2) Macfarland W. Thomas. (1998). Descriptive Statistics and Measures of Central Tendency and Dispersion. Retrieved from http://www.nyx.net/~tmacfarl/STAT_TUT/cent_tnd.ssi Madrigal Demetrius & Mcclain Bryan. (2012). Strengths and Weaknesses of Quantitative and Qualitative Research. Retrieved from http://www.uxmatters.com/mt/archives/2012/09/strengths-and-weaknesses-of-quantitative-and-qualitative-research.php Saline P. Roger. (1999). Measures of Central Tendency. Retrieved from http://courses.wccnet.edu/~palay/math179/c0s3a.htm Trochim M. William. (2006). Levels of Measurement. Retrieved from http://www.socialresearchmethods.net/kb/measlevl.php Read More