Presentation

This paper is being submitted on October 26, 2015 Nicole Engeswick / B312/GEB3124

Section 01 Business Research and Analysis-Online Plus 2015 Fall Quarter.

Background information

The research study deals with the collection of past information about the old alumni of York College. The study will utilize a qualitative research methodology. Therefore, the choice of measurement of scale, conducting the research and accuracy is paramount. This is because measurement errors have to be minimized at all costs. Validity, accuracy, and reliability are integral parts of a research project.

The data collected ought to show these attributes if it is to give meaningful results. The analysis of data to be used requires that the collection process was designed to avoid the possible errors that could compromise the information collected. However, errors in measurement are very difficult to avoid. Data collection would therefore always have some errors. These errors should be accounted for and if possible corrected (Freund and Wilson, 2003). Comment by Nichole Engeswick: Great point

Nominal Scale

The nominal scale will be used for measurement of the statistical data for this research. This scale is easy to use, understand and interpret. This method is used in categorizing the data collected. The classification created does not follow any order or formal structure. There is no much in-depth definition of characteristics to be used to guide the process of grouping the data collected. For instance, in the interview, yes or no are used to collect data for certain research questions. This response is sare a form of nominal measurement. Comment by Nichole Engeswick: not

Nominal scale is mostly used with the non-parametric groups. It is used with modes and cross tabulation. Statistical measurement scales are important when analyzing the types of responses to be received from the interviewees. The questions are qualitative, and, therefore, there is a need to concentrate on the proper analysis of the results.

The measurement scale required for this study is the nominal scale. The nominal scale is utilized when labeling variables or characteristics that do not possess any countable or quantitative values. Nominal measurement scales are paramount in reviewing the views of the respondents of the study, and the main aim is to ensure that the answers are precise to the point. Also, ordinal measurement scale can be utilized in various questions depending on the questions put across in the interview.

Ordinal Scale

This is a type of statistical data measurement scale that seeks to determine the degree of the data variable. It has various forms of responses to allow the research to be varied over a range of values. For example when conducting the research on the level of motivation for learners in a particular class, they would be allowed a varied range of choices such as "highly motivated, moderately motivated, motivated, less motivated, and disillusioned." This would give the researcher more responses from the respondents. Comment by Nichole Engeswick: ?

The data collected would then be classified from the most to the least. Unlike the nominal scale, this one enables the researcher to compare the degree to which the subjects of the study possess the dependent variable. This measurement scale will be used in this research to determine the various experiences that the respondents had in the college. For the institution to be able to pick individuals who would be good ambassadors to the public, it has to establish the level of approval that these alumni have of the college.

Those who hold the school in high regard will find it easy to talk to others about it and to invite them to come and experiences the same. Furthermore, the alumni who rate the programs offered by the school would not think twice to use their connections to refer or recommend the graduates to the employers or to provide them with opportunities at their companies. In appreciation, they will also get involved in the college's funds drive and donate towards the betterment of the institution.

The criteria needed for effective evaluation using this scale, including accounting for errors. In measuring a qualitative research variable, the classical errors in measurement assumes that the errors found in any of the utilized variables in the set of data are not dependent of the true variables used. There is a criterion that has to be followed for adequately evaluating the nominal scale to ensure that the responses given are analyzed effectively (Chiang, 2003).

Select at least two measurement errors that may be a concern using the chosen scales

Measurement errors are always inevitable, and, therefore, they come about while conducting the experiments. The limitations can cause lots of misguided analyzes and should be taken care of. The limitation of nominal measurement is that it has only two values. It does not factor in the values of the measurement variables that fall in between. It cannot be used with continuous data.

Therefore, the data collected has to be rounded off to the nearest whole number. Therefore, the actual value of the measurement variable is lost. The error associated with ordinal data measurement method is the imbalance in the interval. For the method, the difference between the adjacent values in a scale is not necessarily of the equal interval on the underlying scale thus giving rise to the errors in measurement. Comment by Nichole Engeswick: ? Please refer to page 250 for more information on nominal scales. For example a nominal scale would have two or more categories such as are you male or female. In this example you can only be one or the other? I am a little confused on the “rounded off to the nearest whole number”? Comment by Nichole Engeswick: In regards to interval scales they do have a scaled distance between answers. Are you referring to unbalanced and balanced measurement scales?

These errors are to be corrected before the analysis of the results to avoid misinterpretation of having results that do not give an accurate reflection of the variables that is being measured in the study. If not done properly, the outcome may be wrong dealing to inaccurate information.

References

Freund. R. J, Wilson. J. W (2003) Statistical Methods. Academic Press -Mathematics

Chiang. L. C (2003) Statistical Methods of Analysis. World Scientific-Business & Economics

70 - Stephanie – Great job incorporating multiple measurement scales into your research to increase your research data. I think there is a little confusion on the errors. On page 256 the text discusses errors that can occur due to the different measurement scales i.e. the respondent. The next part of the assignment was to state how you would minimize or account for those errors. For example you could have respondent error if the questions were too complex to understand; how you could minimize this error is to design questions that are easy to understand. Keep up the hard work.