Chapter+1

Some thoughts for Chapter 1. Regardless of how well you understand the homework, know these ideas and you are in great shape.

The __individual__ is the thing we want to study, the __variable__ is the specific characteristic of the individual we are interested in.
 * //1. Don't confuse individuals with variables.//**

//**2. Understand the relationship between quantitative and qualitative variables.**// __Quantitative__ data deals with numerical information that can be added, subtracted, etc. together in some meaningful way. For example an individual's age. __Qualitative__ data deals with attributes of an individual such as eye color. So the "trap" is an individual's phone number. The response is numerical but you can't add phone numbers together in a meaningful way, thus phone numbers are qualitative.

Data measurement at the __interval level__ involves numerical responses that can be put in order with meaningful differences between values, however at this level zero simply represents a place on the scale. Temperature is a good example. __Ratio level__ measurements work just like interval level, except that zero means zero. Height is a good example.
 * //3. Know the difference between interval and ratio levels of data measurement.//**

There are many sampling methods you can use to collect data. A simple random sample is the best way to avoid nonsampling error. In a __simple random sample (srs)__ everyone in your target population must have the same possibility of being surveyed. For example, a srs of this class could be performed by placing everyone's name in a hat, then blindly choosing 10 names. Those people would then be surveyed and the data collected would be used to "generalize" the entire class. Keep in mind, no sampling method is perfect, but the srs is the best one we have.
 * //4. Simple random samples are the best.//**

Statistics is as much an art as it is a science. Sometimes, despite our best effort, we choose samples that do not represent our populations. This is called __sampling error__ however it is not a mistake...bad luck perhaps, but definitely not a mistake. On the other hand, __nonsampling error__ belongs to the statistician and usually involves poor sampling technique, poor data collection or the dreaded bias.
 * //5. Sampling error is a way of life.//**

There are graduate-level classes specifically dedicated to the idea of designing surveys (experiments). But here are the basics.
 * //6. Experimental design is complicated. (but very relevant to you nursing majors)//**


 * Surveys (data collection) are done through observation or experiment. __Observation__ involves choosing a variable about a sample and watching for an occurrence of the variable. __Experiment__ involves applying a treatment (something new) to a sample then watching for a change in a variable.


 * __Control groups__ represent a portion of the sample who are not given the treatment and serve as the base group for measuring the effect of the treatment.


 * __Bias__ is the leading cause of death for surveys. There is a nice table on page 24 of the 9th edition that gives some examples of the various types of bias. My personal favorite is the "hidden bias" of question design. For example, Stephen Colbert often asks his more liberal guests the following, "Was George W. Bush a great president or was he the greatest president?"