Survey design


Contextual forces, or “climates of opinion,” can influence the actions and expressions of group members.
Such social norms can be classified into two main categories,

descriptive and injunctive.

Descriptive norms are informational and describe the way things are within a given social setting,


injunctive norms possess a sanctioning function and prescribe the way things ought to be.

Individuals who violate injunctive norms (i.e. engage in forbidden behavior) run the risk of alienating themselves from those around them.

By making use of interactions with:

1-Important reference groups


2- exposure to available media,

individuals are able to get a sense of what is socially acceptable when expressing political views and opinions.

This iterative process establishes the normative environment surrounding
opinion expression—it is within this climate that individuals may feel
more or less inclined to express their own view.

Respondents’ perceptions of congruity between their own opinion and the perceived
opinion of a given reference group can either encourage or dissuade
opinion expression.


Individual perceptions of this climate are an important and quantifiable aspect of understanding the impact that normative environments can have on individual behavior and expression.

Aggregated perceptions of public opinion constitute a “social barometer” that quantifies both

1-the extremity and

2- the amount of agreement among actors as related to these prevailing social forces.

Researchers interested in capturing the normative aspects of public opinion should
account for the following opinion characteristics:

(a) the valence and the strength of individual opinion (e.g. Do you approve or disapprove of X?, and To what degree?); and

(b) perceptions of the valence and strength of the group opinion (e.g. key reference groups, such as members of your community, residents of this state).

Within a survey context, questionnaires need to be geared toward respondents’ perceptions of the climate of opinion and can include questions such as,

In your judgment, what would the reaction be if someone expressed strong support for Candidate X during the course of a conversation among people in your neighborhood: Very positive, somewhat positive, neutral, somewhat negative, or very negative?

While survey items such as these tap into individual perceptions of the context in which opinions are expressed, they also allow for simple (experimental) manipulations of key variables within the question (e.g. substituting candidates or discussion topics, level of support or opposition, and reference groups named within the question).

With this response data in hand, Jay Jackson’s Return Potential Model (RPM) provides one example for quantifying the normative opinion climate. Normative “intensity” can be evaluated using the RPM by calculating the mean group deviation from a midpoint or neutral response option. This measure specifies the approval or disapproval associated with an opinion as well as the extremity or strength with which a norm is held (e.g. slight approval versus strong approval of expressing a specified opinion). It is quantified by using the following equation:

where intensity (I) is the mean deviation from the researcher’s specified
midpoint value (m).

It is important to note that intensity is a
bidirectional concept that can apply to strong disapproval as well as
strong approval. In other words, the degree to which a certain opinion
norm is characterized by negative intensity (i.e. an opinion is
strongly opposed) is related to the probable level of social sanctions
for expressing that opinion. For high-intensity norms, violation will
most likely result in some form of social isolation. On the other hand,
expressing opinions associated with low-intensity norms may be seen as
odd but will likely bear little social cost.

When measuring the normative opinion climate, it also is important to know
how much agreement exists within a specified group. The RPM model also
captures normative “crystallization,” which quantifies the amount of
consensus associated with a norm. Specifically, this assessment of
crystallization measures the level of agreement regarding what opinions
are appropriate to express. Mathematically, crystallization is
inversely associated with variance among respondents’ normative views
and therefore becomes greater when there is relatively more agreement
about public opinion. To properly derive crystallization (C), the
following equation is used:


 where Ci is the inverse of the deviation from the mean approval rating among all respondents. Highly crystallized opinion norms are well understood and solidified.

Low crystallization is likely to be associated with normative ambiguity. For example, if it is perceived that everyone in a
certain group strongly supports Candidate X, it is safer for group members to speak highly of that candidate without fears of social
sanctions. Because such a normative environment is unambiguous (i.e. highly crystallized), group members are better able to anticipate the reactions of others.

With intensity and crystallization available as quantifications of normative opinion climates, researchers are able to generate predictions related to subsequent opinion expression, willingness to deliberate, and any number of communication behaviors that may fall under the sway of public opinion as a social force.

(Carroll J.Glynn and Michael Huge, 2008, 550)


Glynn C. J.
Public opinion as a normative opinion process. Communication Yearbook
vol. 20 (1997) pp. 157–183.

Jackson J.
M. Normative power and conflict potential. Sociological Methods and
Research vol. 4 (1975) pp. 237–263.

Price V. ,
Nir L. , and Capella J. N. Normative and informational influences in
online political discussions. Communication Theory vol. 16 (2006) pp.

Rimal R. N.
and Real K. How behaviors are influenced by perceived norms.
Communication Research vol. 52 (2005) no. (3) pp. 389–114.


Check list:

Identify the constructs.

Identify the variable that you want to measure.

Make sure that the question is measuring the a variable that has the essential qualities of the construct.

Ask one question at a time.

use simple and familiar words

use specific and concerete words

use as few words as possible

use complete sentences

use simple sentence structure.

Put the positive choice first for faster response. (p.146)

Put non-substantive options at the end of answer list. (p.148)

have equal comparisons (p.150)

Are the answers to a question mutually exclusive.

Are the answers to a question collectively exhaustive.

Ideally it is better two versions od the survey in them in which the answer
orders are reverse but consistant within each.

state both positive and negative sides in the question. For example( do you
favor or oppose)

Do you agree or disagree (Dillman et al, 2008, p.119)

Special attention has been paid to identifying if the participant was Given the choice if they actually can be Unsure,

Did not have enough information, Or have thought enough to answer a question (moore,p.144)

Measure the intensity of the opinion If he was upset if not If the issue is important for him Will affect his behavior or attitude (Moore, p. 146)




All, Always, Every, Everybody, Everything: They may be generally in agreement with a proposition, but nevertheless hesitate to accept the extreme idea of all, always,each, every, never, nobody, only, none, or sure. It is correct, of course, to use an all-inclusive word if it correctly states the alternative.

Any, Anybody: The trouble with “any” is
that it may mean “every,” “some,” or “one only” in the same sentence or
question, depending on the way you look at it. Another difficulty with
“any” is that when used in either the “every”or the “not any”context it
becomes as much a deadgiveawayword as are“every”and “none.” : Words with the “any” root
are subject to the same trouble as “any” itself. “Anybody” can mean
everybody or someone person.

Bad: Experienceindicates that
peoplearegenerally less willing to criticize than they areto praise.
Because itis

to get them to statetheir negativeviews,sometimesthe critical side
needs to be softened.For example,

asking What things are good about your job?, we may be wise not to
apply the “bad” stigma but to ask

What things are not so good about it?

Best: Few people do the best they can, for example.

Could:   “Could” means that something can be done.
It should not be confused  with “should” or“might.”

Should: It should not, could not, might not be used as though synonymous.

Might: “might” refers to a probability that something will be done.

Whereas “could” refers to whether something can bed one,


Country: What is meant by this word—the nation as a whole, or rural areas?

Daily:Which is intended—five days a week,or all seven?

Dinner: Dinner, the main meal of the day, comes at noon for some families whereas in some areas, it is
the evening meal. The question should not assume that it is either
the one or the other.

Ever:  “Ever” is such a longtime and so inclusive that it makes it seem likely that something may have happened during the time period of “ever.”
Have you ever seen Seinfeld? “Yes—I suppose I must have at sometime or other.”

Fair: When used in the sense of“just”or “reasonable”         it can be taken to mean “average.”

Few: We cannot assume that this word has definite limits. One person’s few is another’s several.

Government: It is sometimes used as a definite word meaning the U.S. federal government; sometimes as
an inclusive term for federal, state, and local government; sometimes as an abstract idea; and sometimes as the party in power
as distinct from the opposition party. The trouble is that the respondent does not always know which“government”is meant.

It, Its,That, These, These: : These words necessarily refer to some antecedent, and it is best to repeat the full antecedentdent

Just: A word with conflicting meanings. “Just as much,” for example, may mean “only” as much or “fully” as much.

Know: Knowing varies greatly in degree, from mere recognition to full information. Some respondents may hesitate to say they know  Something when they don’t know it for sure or completely. A person may know a song without knowing the words.

This word is usually used as an alternative to “more,” where it may
cause a minor problem. The phrase

“more or less” has a special meaning all its own in which some respondents do
not see an alternative. Thus,

they may simply answer“ yes,more or less” to a question such as

Compared with a year ago, are you more or less happy in your job?    🙂

The solution to this problem is to break up the “more or less” expression
by introducing an extra word or so or

to reverse the two:

Compared with a year ago, are you more happy or less happy in your job?

Compared with a year ago, are you less or more happy in your job?

Like: This word is a problem only because it is sometimes used to introduce an example.

The problem with bringing an example into a question is that the respondent’s attention may be
directed toward the particular
example and away from the general issue it is meant to illustrate. The choice
of an example can affect the
answers to
the question – in fact,it maymateriallychangethe question, asin
thesetwo examples:

thinkthat leafyvegetableslikespinach should be in the daily diet?

thinkthat leafyvegetableslikelettuceshould bein thedailydiet?

do not like spinach, theywould answer“no”to thefirst question.

More: This wordhas moreor less
beendiscussed underthe word“less.”Itis aproblem foranotherreasonalso:

“more” is used in the comparative sense, it is usually advisable to
indicate the basis for comparison—


finding question wording more complicated?

Most: This
wordcanintroducetrickydoublethoughts asshownbythis question:

Where would you be
doing the most useful work?

is meant—the most work that is useful or work
that is the most useful?

Much: “Much” is an indefinite
word. The “how much” type of question leads to unnecessarily wide

response—questions should be worded such that it is clear that the
responses are expressed in specific terms

percents,fractions,and othermeasures.

Never: A deadgiveawayword.

Nobody: This is
yetanotherdeadgiveawayword.Nobodycanuse “nobody”with impunity.

None: This also canbea

Now: For a word that appears
reasonably clear, “now” can be almost too definite in the sense of
“right this

leadingto situations likethis:

work areyou doing now?


This definite-sounding word is not always so definite. Some homeowners
think that they will not own

homes until they pay off the mortgage. Some stockholders have no
feeling of owning part of their


Possible: An alternativethat uses
“possible”in the ultimatesense(“asmuch as possible”)is adeadgiveaway.

This word is quite frequently misused. “Quite a little,” for example,
has no sensible meaning. If in a

the word “entirely” can be substituted for “quite” without changing the
meaning, then “quite” is being

However,in such properuse,“quite” maybecomea deadgiveawayword.

See, Seen: These words are sometimes used in the sense of visiting
someone, but they may be interpreted


you seeyour dentist last?

the golfcourse.”

Service: Here is another indefinite
word. Try, for example, to put down exactly what you mean when you

the“service”of the electricutilitycompany.

Such: Beware of this word because
it is often used to introduce examples. When we discussed “like” we

out that the particular example may supplant the general issue in the
minds of respondents. “Would you use such a product to clean your
carpet?” may deflect the respondent’s attention from this product to
another one.


Today: This may be interpreted too literally, just as “now”maybe.

Are farmers getting a fair price for milk today?

“Do you mean right today? You know the price dropped this morning.”

Trip: This wordneedsto bequalified—”one-waytrip” or“round-trip,” forexample.

Where: Theframesof referencein answersto a “where”question mayvarygreatly.

Where did you read that?

“In the New York Times.”

front of thefire.”


You: How many computers did you
repair last month?This question seemed to work all right until one
repair person in a large shop countered with, “Whom do youmean, me or
the whole shop?” Sometimes “you” needs the emphasis of “you yourself,” and sometimes it just

the rightwordto use,as in the aforementionedsituation.


Glynn, C., & Huge, M. Opinion Norms. In Paul J. Lavrakas (Ed.), (2008) Encyclopedia of Survey Research Methods. (pp. 550-551). Sage Publications, Inc. doi:10.4135/9781412963947.n351


Lavrakas, Paul J. 2008. Encyclopedia of Survey Research Methods. 1 edition. Thousand Oaks, Calif: SAGE Publications, Inc.


Australian Election Study, 2001

Paul, H. (2005). Strengthening Canadian Democracy. IRPP. introducing direct democraccanaday in canada MatthewMendelsohn

Australian election studies, 2001

Haslam, S A., & Reicher, S. D. (2012). When prisoners take over the prison: a
social psychology of resistance. Personality and social psychology
review: an official journal of the Society for Personality and Social
Psychology, Inc, 16(2), 154–179. doi:10.1177/1088868311419864

I will use stratification for activist or not in my sampling based on backgroud data and actual answer. I will use telephone random generator. I will create an email database and randomly choose from it I will use other email sources I will use face to face I will use avalanche i will keep the track of method of selection The following ratios need to be corrected if not automatically proportional gende age race-ethnicity income education



Test Instruments

Political Compass