Rapple+Journal+1

Back to Lisa's Journal Go to Journal 2 When a group has been identified for study, all the members of the group are called the "**population**". When a subsection of this population is created it is called a "**sample**". It is desirable for the sample to be representative, in other words, have the same proportions as the larger population it represents. This is called **EPSEM** - equal probability selection method. Just as sampling size is representative of population, so is statistic representative of parameter. See also sampling
 * Definitions/distinctions of populations & samples; Attributes and variables; levels of measurement**

//The Pew Study: Social Side of the Internet Survey 2010// //The population is all adults in the continental United States who have access to landline or cellular phone. The sample frame (??) is all telephone listings. The method used for sampling was two-fold. For the landline surveys probability sampling was used. For cellular telephones that had no listings, a systematic sampling method was used.//

When evaluating study data Deveaux suggests answering the "w" questions. Who is usually identified in 1st column, each row is an individual case. So WHO does each data table/row refer to? The "who" may be; respondents (surveys), subjects or participants (experiments), experimental unit (inanimate subjects), or observations (vague term for "who"). The variables will answer the question of "what", the variables identify what has been measured. These are the variables or characteristics about the individuals. The last question to be answered is WHY you are examining the data. Deveaux describes this process as a habit of data analysis. It is a systematic way of putting the data collected for the study into context. Identifying qualitative and quantitative variables means looking at what the units of measure are. Also it is important to differentiate if they are simply serving as labels or categories. Variables fall into three types: **Nominal** variables are labels. Each datum is given a category membership such as yes/no, on/off, is/isn't. **Ordinal** variables give a score related to the other scores on a range. And finally **scale** variables place data on an independent scale with units of measure. Each datum stands alone, independent of the other datum.Read more Quantitative data can be further broken into subgroups of discrete and continuous. Discrete is a counted number of occurrances, whereas continous is measured on a scale (i.e., inches on a ruler, etc).
 * Independent** **variables** are characteristics or data that can not be changed by other factors or variables. **Dependent** **variables** are characteristics or measurements that will change when other factors or variables change. For example, someone's age is an independent variable. Nothing can act upon a person and change their age (that we know of). A dependent variable may be a behavior that will change as circumstances for a person change.

//The Pew Study had independent and dependent variables some of which I identified and modified for my study for this course: view my variables//

With quantitative variables there are a couple of cautionary points to evaluate. If the units are in ranges they may be categorical, ex: age range versus age value. Also, with ordinal variables when you are measuring perceived value, be very cautious about treating these as pure quantitative (ex: numbered rating scales/likerts) data. In differentiating qualitative and quantitative COUNTS, where you are counting the number of something, consider that; # in a category is categorical (qualitative), whereas amount/measure of something is quantitative. Be sure to RECOGNIZE IDENTIFIER VARIABLES - recognize these so you don't try to count them. Identifier variables have categories of "one". They are used to; provide unique labels, provide confidentiality for the "one", to combine data from different sources. Identifier variables can be categorical or quantitative. For example, Year can be a category or it can be a measure of timespan.

Good! You have demonstrated a good understanding of sampling and variable measurement!