Class Notes (836,068)
COMM 2002 (65)
Lecture

# COMM2001 October 21, 2013.docx Premium

4 Pages
95 Views

School
Department
Communication Studies
Course
COMM 2002
Professor
Heather Pyman
Semester
Fall

Description
Univariate data analysis • Surveys and the construction of different types of questions • How to describe different distributions and that data that we are looking at o Now we are moving onto data analysis Getting Started • Think about data analysis at an early stage in the research process • Decisions about methods and sample size affect the kinds of analysis you can do • Level of measurement critical to analysis • Data analysis that is possible that is important to the research analysis • For us, what we need to take away from this is that all the data that we are looking at, each question is measured at a different type of measurement • Run on a different type of measurement that someone else collects for us • Variables can be constructed to varying levels of complexity Levels of measurement • Nominal o A set of distinct categories • Ordinal o Response categories with order to them • Interval o Reponses categories have order plus equal measurable distance between categories • Set of names with no numerical value to them • When we think about political parties for example, we will have a set of distinct categories • We have to assign a numerical value to each of these labels • SPSS does not read words, only looks at values that are assigned to those categories • Important thing to note that SPSS does not know whether there is a real value to the value of one, two, etc. • Could just as easily start numbering with no difference to information that SPSS is looking at -> Category and value has no meaning • Basis of what can and cannot be done that are measured at the nominal number • Nominal = name • If we have a variable that is just a list of things -> political parties, environmental groups, baseball names, etc. just a numerical value • Move from nominal measure to ordinal measures -> go to measurement of scientists • Order of categories from low to high or low to high, the points of the scale -> each point sounds like some more or some less • Think of agreements -> strongly agree, neither, disagree, etc. • Can also have options such as always, sometimes, often, never • Or not important, somewhat important, very important • Assign values to everything, still 1, 2, 3, 4 (not applicable) • No way to quantify the agreements, etc. They are not units, have to know when we have real meaning and when you don't • All we know with ordinal measures is whether the scale is low to high or high to low, cannot quantify it o There is order to it, but no real mathematical meaning o Final measure is called interval measure • Real mathematical meaning to the points on the stage • Look at people's ages, or at a thermometer, or a ruler, we look at measures • With this data, it is difficult to look at it, so we do collapsible data, put interval level into different variables with order to it -> ex. 18-19-20-21-22 etc. but can put into 18-24,25-34, 35-44 • Important to note that at the interval level with real meaning in the scale is that it is different Examples: • What is the temperature in this room? _______________Co • On a scale of hot to cold, how would you describe the temperature in this room? Is it… Hot Warm Cool Cold • Which of the following describes the temperature in the room? Stuffy Chilly Hot Pleasant • We know that scales go from high to low -> can have values of different numbers • When SPSS is doing the analysis, it assumes that you know what the data is • Different correlation coefficients, we know if the scales are from high to low or low to high Why do we care? • Advantages of Higher Level measurement o Precision o Ability to use better statistical measures • Disadvantages o Difficult to look at distributions o Asking questions about education or support • Ex. People who are more educated are more likely to support Harper • We would have the options • Look at how variabil
More Less

Related notes for COMM 2002
Me

OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Join to view

OR

By registering, I agree to the Terms and Privacy Policies
Just a few more details

So we can recommend you notes for your school.