DISCLAIMER: These notes are based upon my readings and studies of the given material, provided by our instructor. DO NOT BELIEVE that these notes will directly correlate with a passing grade on the next midterm. These notes reflect my understanding of the material and therefore are subject to possible (albeit minimal) but possible error. This is not a substitution for reading the assigned chapters and articles, or attending class. With that said enjoy, and good luck on Friday.
CHAPTER 7
Quota Sampling (nonprobability) units are selected into a sample on the basis of prespecified characteristics.
-the total sample will have the same distribution of characteristics assumed to exist in the population being studied.
Simple random sampling (probability) the units composing a population are assigned numbers. A set of random numbers is then generated, and the units having those numbers are included in the sample.
CONS: “Not feasible” [?]
Not the most accurate method available.
Systematic sampling (probability) every “kth” unit is selected for inclusion in the sample.
EX: every 25th student in the college directory.
[K=sum of population/sample size = sampling interval.]
PROS: slightly more accurate than simple random sampling
CONS: Arrangement of elements within the list might make you select a sample that does not reflect the population.
Stratified sampling – modifies simple random/systematic sampling – ensures more representativeness by decreasing sampling error.
PROS: Organizes the population into homogenous subjects and selects the appropriate number of elements from each.
Offers proper representation of the stratification variables
Multistage cluster sampling – natural groups (clusters) are sampled initially with the members of each selected group being subsampled afterward
-involves repetition of listing and sampling
PROS: great for stratified populations, can break down your population far better
CONS: subject to two sampling errors
Purposive sampling (nonprobability) the units to be observed are selected on the basis of the researcher’s judgment about which ones will be the most useful or representative.
PROS: Good for deviant samples
Good as a pretest
Good for insights on a sample
CONS: Weak for measuring the totality of your sample
Snowball sampling (nonprobability) – often used in field research each unit interviewed is asked to suggest another unit to be interviewed
CONS: can lack representativeness
PROS: Great for people who are hard to locate
EX: Studying the homeless
Confidence Interval – range of values in which a population parameter is estimated to lie
± Z ( Ѕ . )
√N – 1
Ci = sample mean ± confidence level (standard deviation ÷ √ Sample size – 1)
Confidence level – The estimated probability that a population parameter lies within a given confidence interval.
3 confidence levels
– 90% (rarely) i.e. there is a 90% chance that the sampling statistic equals the population parameter and a 10% chance that it does not. [1.65]
– 95% i.e. there is a 95% chance that the sampling statistic equals the population parameter and a 5% chance that it does not. [1.96]
– 99% i.e. there is a 99% chance that the sampling statistic equals the population parameter and a 1% chance that it does not.[2.58]
What is a standard deviation?
Standard Deviation = answers the question how disperse are a set of values around a sample mean.
Mean is calculated by: Sum of individual values reported
Sample Size
Standard Deviation: √(square root) ∑ ( Individual value reported – sample average)2
Sample Size
Chapt 8:
Classical Experimental Design: most conventional –
needs: Independent and dependent variable(s)
Pretesting and post testing
experimental and control groups
Examines effect of IV (independent variable) on DV (dependent variable)
IV – experimental stimulus
2 attributes (present/not present)
Both IV and DV must be operationally defined before experiment begins
Experimental group – the group who is administered the stimulus
Control group – no stimulus – has characteristics very similar to experimental group
Chapt 9:
Open-ended questions – respondent gives his/her own answers (most qualitative studies use this in-depth method)
Closed ended questions – respondent selects a given answer, provided by researcher.
Guidelines for Asking Questions :
• Make items clear
• No ambiguous questions
• Avoid double barreled questions (2 ?’s in 1)
• Respondents must be competent to answer.
• Respondents must be willing to answer.
• Questions should be relevant.
• Short items are best.
(If question to long, people will move on, miss read it, or get bored)
• Avoid negative items.
(avoid using the word ‘not’)
• Avoid biased items and terms.
(welfare vs poor)
Guidelines for Questionnaire Construction
• One question per line; avoid clutter; clearly indicate how and where respondents should answer.
• Format matrix questions so they are easily answered.
• Use contingency questions when necessary.
• Be aware of issues with ordering items
• Include instructions for the questionnaire.
• Pretest all or part of the questionnaire.
