Lecture Notes on Statistical Methods | STA 100, Study notes of Data Analysis & Statistical Methods

Download Lecture Notes on Statistical Methods | STA 100 and more Study notes Data Analysis & Statistical Methods in PDF only on Docsity! Prof. Thistleton STA100 Statistical Methods Lecture 1 Text Sections: Chapter 1 Basic Goal of the Course: Derive useful information from data, i.e. try to answer questions we find interesting. Example: Suppose that you have heard that the rate of “turnover” for Nurse’s Aides who work in Hospitals is very high and you’d like to try to find out why. The first thing you would do is to begin to read the literature about turnover in various occupations and to try to find articles that are pertinent to your particular setting. For example, Bower’s has an article (Bowers, Esmond, & Jacobson, 2003) that you might find interesting. Also, you see that Castle (Castle, Engberg, Anderson, & Men, 2007) has a survey form that you would like to administer. The questions comprising this survey are included here for reference: 1. Rate how much you enjoy working with residents 2. Rate how your role influences the lives of residents 3. Rate your closeness to residents and families 4. Rate the care given to residents 5. Rate the effect you have on residents’ lives 6. Rate whether your skills are adequate for the job 7. Rate the training you have had to perform your job 8. Rate chances you have for more training 9. Rate the people you work with 10. Rate whether you feel part of a team effort 11. Rate cooperation among staff 12. Rate the support you get when doing your job 13. Rate the chances you have to talk about your concerns 14. Rate the demands residents and family place on you 15. Rate your workload 16. Rate your work schedule 17. Rate the amount of time you have to do your job 18. Rate how fairly you are paid 19. Rate your chances for further advancement 20. Rate your overall satisfaction with your job 21. Would you recommend working at this facility to a friend? Each of these questions may be answered with a number from 1 to 10. (All questions used a 10-point visual analogue rating format scale ranging from 1 (very poor) to 10 (excellent).) So, suppose you administer this survey to 200 Nurse’s Aides at your facility. How do we begin to make sense of the answers? Prof. Thistleton STA100 Statistical Methods Lecture 1 Typical “Statistical” Activities understand what it is you would like to know, i.e. construct a precise research question design/decide upon data collection methods and data analysis methods collect data analyze data interpret data answer your questions Some Basic Terms Objects of Study (also called cases, individuals, etc.). In the above example we are studying Nurse’s Aides in general, the Nurse’s Aides at our facility in particular, and finally we obtain data for those individuals willing to participate in our study. Variables: characteristics of the objects of study. For each of the study participants we have responses to 21 questions, so we so far have 21 variables associated with each object of study. You will also probably gather some basic demographic information such as o Age o Gender o Years of Education o Weight (not really likely, but this sets us up for the next part of the discussion) o Etc. Each of these will provide you with another variable to study. Notice that these variables describe different types of information about your objects of study. Samples and Populations: Usually we are collecting data that we have at hand to try to generalize to a larger group. Following the above example, we might administer our surveys to workers in a few hospitals where we have access to the Nurse’s Aides but it is more interesting to be able to talk about Nurse’s Aides in the USA. Learning how to generalize from the group we have at hand (the sample) to the population of interest is a major goal of this course. Levels of Measurement Our variables are used to acquire different types of information and must be handled appropriately. For example, it makes sense to talk about the average age of your study participants, but what about average gender? This leads us to carefully consider the level of measurement for a particular variable. Our text tells us about the following: 1. Nominal Data: (in name only) categories e.g. good/bad/ugly, male/female 2. Ordinal Data: (ordering introduced) relative comparisons are possible e.g. low/medium/high 3. Interval Data: ordinal with equal distances between rankings e.g. degrees Fahrenheit 4. Ratio Data: interval data in which arithmetic division is possible. e.g. weights This last one is a little tricky. The basic difference between Interval Data and Ordinal Data is that division makes sense with ratio data: 5 pounds is really half of 10 pounds, but is 5 degrees Fahrenheit really half as warm as 10 degrees?