PSYB01H3 Study Guide - Final Guide: Skewness, Compact Space, Frequency Distribution

75 views5 pages
26 Jan 2013
Surveys and Questionnaires usually large scale, want to get information on whole population,
sampling strategies are important
-Questionnaire instrument used to conduct a survey concerned with individual responses
Steps in Questionnaire Development
1. List variables: background, dependent, independent
2. Operationalize variables
3. Decide how data will be analyzed “types of questions”
4. Develop wording for questions
5. Write proposed questions on index cards to facilitate editing and rearranging order
6. Pre-test the questionnaire- do they understand the questionnaire? Instructions? How long does
it take?
7. Shorten list, refine questions
Presence Absence Questions respondents check off which items in a list do or do not apply to them
less commonly used (Yes/No)
Single Choice Question ask respondents which one category applies (What year are you in university?)
Likert Type Questions respondents indicate how much they agree or disagree with a statement
(Strongly disagree …. Strongly agree)
Graphic Rating and Non Verbal Cues for people with low literacy skills ( …. )
Rank Ordering Questions should be avoided or minimized because it takes time to answer
Semantic Differential Scales used for measuring the meaning of concepts (unreliable … reliable)
Open Ended Questions pros: permits details, clarification, unanticipated answers, reveals logic behind
respondents’ response – cons: bias towards educated, irrelevant answers possible, coding and statistical
analysis difficult, generalization or comparison difficult
Caution: time consuming to answer and code, generate responses that are inconsistent, likely to be left
blank, use if little is known
Pitfalls to avoid double barreled questions, hidden assumptions, question and answer don’t match,
leniency bias, leading questions, alternative meanings, questions with jargon, imprecise questions,
vague/ambiguous questions, social desirability, threatening questions
Kinsey Technique emphasize on continuing of the graduations between always and never
Populations and Samples inferential statistics based on inferring from a sample to a population
Gerenalizability ability to infer population characteristics based on the sample
Sampling Error difference between sample and population characteristics, reducing sampling error is
the goal of any sampling technique, as sample size increases, sampling error decreases
Process of Sampling Selection Naming of population (can be people, organizations, events), restrict by
age, geography, etc. Determining Population Size depends on total population and desired confidence
Employing Appropriate Sampling Strategies Probability Sampling likelihood of any member of the
population being selected is known and equal
Non Probability Sampling 0 likelihood of any member of the population being selected is unknown
Simply Random Sampling each member of population has an equal and independent chance of being
chosen, sample should be very representative of the population, ideal in theory but difficult in practice
Stratified Sampling goal is to select a sample that is representative of population characteristics of
interest are identified (eg. Gender), individuals in population are listed separately according to their
classification (eg. Females, males), proportional representation of each chart is determined (40% female,
60% male) a random sample is selected that reflects the proportions in population (4 females, 6 males)
Cluster Sampling identifying clusters of the population and randomly selection from them, the whole
cluster is then used, units must be homogenous in order to avoid bias
Unlock document

This preview shows pages 1-2 of the document.
Unlock all 5 pages and 3 million more documents.

Already have an account? Log in
Non probability Sampling non random sampling refers to strategic requests for ‘volunteers’ the use
of informants that snowball or hand picking respondents selecting a sample on basis of convenience
can threaten a study’s credibility
Convenience and Snowball Sampling captive or easily sampled population not random, weak
Quota Sampling proportional stratified sampling is desired, but not possible
Quasi Experiment don’t meet requirements necessary for controlling the influence of extraneous
variables, usually random assignment is a missing piece
Features of Quasi Experiment: matching instead of randomization, time series analysis, unit of analysis
often something different than people
Threats to Internal Validity: history some event affects the study out come, maturation natural
change in subjects overtime, instrument change not using the same measurement tools or methods
with each subject
Testing/Repeated testing fact of having been tested, practice effects prior exposure to a
measurement, fatigue prior participation tires the participants
Mortality participants dropping out
Regression to the mean high/low measurements tend to be followed by measurements that are closer
to the group mean
Time Series Analyses useful when you cannot randomize participants and where it is possible to obtain
a series of assessments of the dependent variable at pre treatment and post treatment
Single Case Research Designs use only one case or one group to investigate a specific phenomenon
not the same as case study, uses time series design
Advantage: avoid problems with group mean, can examine participants from hard to find population,
can deal explicitly with individual (not group) behavior, results are easy to interpret (no stats), can focus
on helping few participants
Disadvantage: hard to demonstrate causality, no controls in most cases, lack of statistics, can’t really
look at interaction effects, counterbalancing is problematic, problem of external validity
Take multiple pre test and post test measures to overcome limitations of a one group (case), still a quasi
experimental design, cannot exclude natural history as an explanation for an effect
Phantom pain attention diversion used to assess pain control (A-B design)
Cocaine Abstinence (A-B-A design) escalating reinforcement for cocaine free urine samples/ Baseline
reinforcement, then withdrew reinforcement
Problems with A-B-A Design end up with baseline condition, need to add another treatment phase
A-B design: not all treatment related behaviors may be reversible
Multiple Baseline Design useful in testing for a treatment effect when you believe that effect is
Baseline Data are collected on 2 or more behaviors for same individual, same behaviors for 2 or more
individuals, same behaviors across 2 or more situations for same individuals
Data Analysis all pieces of the research plan must connect research questions, measures, data
collection and analysis
Scales to consider nominal: in name only categories are only labels: gender, ethnicity, marital status
Ordinal: response options in an order that makes sense, but difference between options are
meaningless (rank of priority, 1=best, 5=worst)
Interval: difference between a score of 50 and 51. But, someone with a score of 40 is not half as anxious
as a person with a score of 20. Ratio: same as interval but has true zero and ratios make sense/ eg.
Weight, height, number of children
Descriptive Statistics: summarize your data set for the benefit of others characterization of data
sport statistics are descriptive, deal with only data in your sample
Unlock document

This preview shows pages 1-2 of the document.
Unlock all 5 pages and 3 million more documents.

Already have an account? Log in

Get OneClass Grade+

Unlimited access to all notes and study guides.

Grade+All Inclusive
$10 USD/m
You will be charged $120 USD upfront and auto renewed at the end of each cycle. You may cancel anytime under Payment Settings. For more information, see our Terms and Privacy.
Payments are encrypted using 256-bit SSL. Powered by Stripe.