PHIL 1200 Lecture Notes - Lecture 7: Causal Reasoning
Polling
Surveys used to determine what proportion of people in a group have a given property. Polls are a kind
of generalization.
Sampling errors:
Size: sample needs to be large enough
Representative: easy to make mistakes here
Ideal: random sample
Margin of error
A poll gives us a specific proportion of the Xs in the sample that have property Y
The margin of error tells us what the probable range or proportions of Xs that have Y in the whole
population
Confidence level
Tells us how confident we are in our generalization of the proportion of Xs that are Y
This tells us that if we conduct the poll 20 times, 19 of those times will show that between 47% and 53%
of people would buy electric vehicles.
Measurement errors
Does the sample really have the property the pollsters say?
Easy mistakes to make!
-Vague language
-Timing of the poll
-Who conducted the poll
-who commissioned the poll
-other things: order of the questions; priming
Reporting on polls
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Surveys used to determine what proportion of people in a group have a given property. A poll gives us a specific proportion of the xs in the sample that have property y. The margin of error tells us what the probable range or proportions of xs that have y in the whole population. Tells us how confident we are in our generalization of the proportion of xs that are y. This tells us that if we conduct the poll 20 times, 19 of those times will show that between 47% and 53% of people would buy electric vehicles. Newspaper articles are usually more accurate than headlines. For the best information, go to the source. General causal reasoning: about kinds of events causing other kinds of events. Constant condition: causal factor that must be present if an event is going to be caused to occur. Variable condition: the event we designate as the cause.