INFO 370 - L06 - October 18, 2004
Notes By: Prins, Fortier, Egaas
Format: UTF-8
Retarded ASCII time!!
Left Skewed (Negative)
_____|
__/ |
__/ |
_/___________|
Right Skewed (Positive)
|____
| \___
| \___
|____________\
saxton iddn't massage the data heheehehhe
1 yes
2 no
3 don't know
4 SYSMISSING
NEW SHIT
1 Yes
2 No = 0
3 No = 0
9 SYSMISSING
SYSMISSING No
Frequency Chart (on green handout)
Internet Poll -- Cheap -- Low Response Rate
Telephone Poll -- Expensive -- Highest Response Rate
In-Person -- Expensive -- Highest Response Rate
Confidence Intervals (Confidence level = z score)
99% -> need to be sure - critical medical studies (p: .005, z: 2.576)
95% -> Kinda need to be sure (p: .025, z: 1.960)
90% -> less sure (p: .05, z: 1.645)
--Unicode text below--
n = ((z*σ)/m)² --- Sample size desired for margin of error
Quiz on Wednesday
Descriptive versus Experimental Study
> Descriptive
* To find out characteristics of the populations
* Snapshot of the population
* Does not always have a hypothesis
> Experimental
* Cause and Effect
* Explanatory
* Predictive
* Impose certain controls (Explicitly defined variables)
* ALWAYS HAS A HYPOTHESES
Exploratory verses Descriptive
> Exploratory
* No Hypothesis
* Open ended questions
* Definition of variables usually come while doing the study
* Study something in its natural setting with few or no preconceived ideas
What makes things forms of evaluation
> Has to be something in place
> Set standards for which you judge outcome, an evaluation
Review definitions for levels of measurements
> Nominal
> Ordinal
* Categories could be evenly spaced
* Doesn't make sense to talk about 0.5, 0.4 (no numerical relationship)
> Ration
> Interval
* No absolute zero
* Even interval (numerical relationship)
Participant and Non-Participant Observation
> Participant
* Watched them watching TV and their habits
> Non-Participant
* Asking them about how they watch TV and their habits
Format
> Given various situations then you answer them
> Multiple Choice
> Check all of the slides
SAMPLING
Population
Population: A group of people or objetcts you are interest in studying
Sample: A subset of the total population
Unit: An individual member of the population (case, subject respondent)
Sampling
Sampling: The process of selecting units for observation
Sampling Frame: Source from which the selection was made (e.g., a list)
Variable: A characteristic of a unit
Observation: Noting the value of a variable for a given unit
Sampling Error
> Any difference between the characteristics of a sample and the characteristics of the population from which it was drawn.
Identify the Population
> Define the population concretely
> Examples:
- Adult residents of Seattle
- Official homepages produced by community colleges in the United States
- Phone calls received at the technical support service desk for a software company
Census vs. Sample
> Cencus
- Observe each unit
- An "attempt" to observe the entire population
> Sample
- Observer a sub-group of the population
Types of Sampling Methods
> Probability (random)
> non-probability (non-random)
Random Sampling
> Objective
> Each unit (element) has the same change (probability) of being picked)
> Change determines who is in the sample
Types of random sampling (KNOW THESE)
> Simple random sample (SRS)
> Systematic sampling
> Stratified sampling
> Cluster sampling
Simple Random Samples
1) Obtain a complete sampling frame
2) Give each unit a unique number
3) Decide on the required sample size
4) Select that many numbers from a table of random numbers
5) Select the units which correspond to the randomly chosen numbers
Systematic Sampling
> Sample fraction - divide population y the desired sample size
> Select from the sampling frame according to the sampling fraction
- e.g. sample faction = 1/5 means that we select one person for every five in the population
> Starting point
- First unit (random first place to start)
- Random
Stratified Sampling
> Premise - if a sample is to be representative then proportions for various groups in the sample should be the same as in the population
> Stratifying variable
- characteristic on which we want to ensure correct representation in the sample
> Order sampling frame into groups
Cluster Sampling
> Involves drawing several different samples
> Draw a sample of areas
> Start with large areas then progressively sample smaller ares within the larger
- Divide city in districts - select SRS sample of districts
- Divide sample of districts into blocks - select STS sample of blocks
- Draw list of households in each block - select SRS sample fo households
Benefits
> Minimizes bias of investigator
> High probability that sample generally representative of the population on variable of interest
> Ability to generalize from sample to population using statistical techniques
Non-Random Samples
> Availability
> Quota
> Purposive
> Snowball
Availability Sampling
> When we study only cases that are available to us
> Examples:
- Surveys by TV station
- Running an experiment with students in your university
- Interviewing kids whose parents gave permission
> When?
- Exploratory research
Quota Sampling
> Representation is established at levels defined by the investigator
> Selecting cases though availibility sampling or random sampling
> NOT PROPORTIONAL
Purposive Sample
> Each single case is selected for a purpose, because of the unique position it has.
> Examples:
- Top 5% of students in the department
- Bottom 5% of students in the department
- Most popular search engines
- Web browser with highest market share
Snowball Sampling
> One identifies one person in the population and ask her or him to identify additional people, and so on.
> Examples:
- Members in an online community
- Participants in a secret and exclusive listserv
- Members of the same professional organization
# END #