Class Notes (1,100,000)
CA (620,000)
York (40,000)
PSYC (5,000)
PSYC 2030 (200)

PSYC 2030 Lecture Notes - Likert Scale, Semantic Differential, Behavioural Sciences

Course Code
PSYC 2030
Lisa Fiksenbaum

of 4
Exam- 97 multiple choice
Mostly from textbook
Greater emphasis on last 3 chaps.
2hrs n 15 minutes.
Who conducts survey?
Major tv networks
Statistics Canada
Auto manufacturers
Magazine and trade journals
Used extensively in the behavioural sciences
Gather large amounts of info
Useful way of obtainin info abt ppl’s opinions
Response options and questions can vary
-imp o make sure when writin questions you don’t introduce any baised
-question shud b simple, straight forward, ppl should understand the question
-be specific
-avoid technical terms
-avoid double-barreled questions
- avoid options that are not mutually exclusive
-avoid leading and loaded questions
leading implies that a certain response is desired (most people favour the use of
nuclear energy, what do u think?)
loaded contain emotion laden words (people who discriminate are racist pigs)
-avoid double negatives (never should one not help others)
-never assume anything. (do you agree or disagree with US trade policy toward
-address sensitive topics sensitively.
-responses to questions
-open ended (respondents state their views in their own words) while quite rich
in the info you get, takes much more time and effort in terms of analyzing it
-close ended ( respondents must limit their answers to the choices present)
more structured approach
easy to code, recodrd and analyse results quantitatively.
-info not as rich as the info, not exactly how they feel in a situation or abt that
-Likert scale, 5 point or 7 point. You wanna have half of the items vely and half
+vely phrased.
-semantic differential scale, presenting a set of bipolar adjectives paris such as
good-bad, strong-weak. Each item is scord on a 7 point sclae, usually rangng
from -3to+3. Concepts are rated along 3 along 3 basic dimensions. The problem
is to use adjectives that are appropriate for what you are measuring.
-graphic rating scale, requires a maork along a continuos millimeter linethat is
anchored with descriptions at each end. Sometimes known as a feeling
-non-vebal scale, use emotions to indicate response options. Always best to use an
existing scale than to make up a question of your own as it has been thru that
process several times and been tested.
- question placement plays v.imp role. Can affect responses to them. Should
start out with questions that are easy to answer, but that are also interesting.
Demographic items should not be presented first (wen you’re conducting an
interview, you want to build some trust, so you’ll ask them demographic
info). Questions should move from the general to the specific. Should have
some continuity, related items should be presented together; rather than
jumping from issue to issue.
- Layout of questionnaire, use closed-ended questions. Use a consistent item
format. Don’t crowd questions together. Don’t split questions or options
between pages. It is most likely to be completed easily if easy on the eye,
relevant, logical and as short as possible.
- Responses biases.
Social desirablility, caused by motivation to make a good impression on
Acquiescence response- tendency either to agree/disagree with statements
Likely to manifest when, people lack the skill or motivation to think about the
ansers, a great deal of thought is required to answer. Respondednt are
unsure abt how to respond.
Handling response biases, avoid social desirability by wording questions so
that ach alternative appears socially desirable.
Acquiescence response= balanced measure…reverse scoring.
Administering surveys
- group administration, respondents are brought together to complete
questionnaires, difficult to motivate respondents, no way of preventing ppl
from skipping questions.
- Personal interview, time consuming. Opportunity for interviewer to probe
inadequate responses. Ask for clarificaion. Controls the order fo questions
being asked. Lack of respondent anonymity. Opportunity for interviewer
bias, might be selective in terms of the information.
- Telephone interviews, lower cost and completed more quickly than personal
inteviews. Ppl have the opportunity to hang up if they’re not interested or get
bored. Less opportunity for interviewer bias to affect responses. Avoid
questions that are too long or complex. Participants might not really trust
who you say who you are.
- focus gropus, semi-structured group interview. Determine the group
members’ perceptions of and affective resnses to that topic or expericne.
Usual open ended, b/c they wnt to explore a topic. Gropu size, if too large
could be dominated by a few members. 6-12 is ideal. Adv, may provide
provocative aor/or nsightful info. Disadv, some ppl less comfortable than
others in expressing their opinions publicly. Vulnerable to interviewer bias.
-Checklist for survey use, do the items provide a representative of the beahviour
under study. Thru pilot testing you can check if the questionhas the same meaning
to the respondent, how might the questions be arranges to encourage the
respondents participation.
-advantages, provide an abundace of fata in easy quantifiable from . data can be
typically analyzed with powerful statistics method.
disadvantages, may only reveal superficial feeling about situation. Gives the
researcher evry little control. Response rate frm spme form of survery are low.
Cannt draw causal inferentces.
-Population and sample
-different samplng methods.
-reperesentative sample, closely matches the charateristics of the population to
whom one wishes to generalize the findings.
-Biased sample, one that is not representative od the population, likely the result of
selection bias.
-sampling approaches, non-probability sampling (no guarantee that eah element has
some bachne of being included), Probabability sampling( each person in the
population ha a known chance of being selected)
-probabiliy sampling, investigator must have a complte list of all the people in the
population. Investigatior randomly selects form the frame a sample of a given
number of people, using a table of random number, a computer random number.
-problems, ma be systematic features of population. Procedues may be impractical if
population is large.