Up A Brief Quiz Survey Studies

Chapter 9

Design of Descriptive
Empirical Research
in Communication

OUTLINE CONCEPTS

 

I. Invitations to Empirical
   Research in
   Communication
   --problem statements
     and hypotheses that
     focus on current
     descriptions and
     relationships
     (attempts to:
     ascertain norms;
     establish goals;
     develop methods)
  --data that are
    quantitative; research
    questions often
    addressed by a
    combination of
    methods to
    "triangulate" research
    methods to draw upon
    all available
    information
empirical: means observable; most people use the term to refer to "scientific research"
experiment: a study that introduces a variable and determines its effects while controlling all other variables
II. Survey Research survey: the process of looking at something in its entirety
survey research: empirical study that uses questionnaires or interviews to discover descriptive characteristics of phenomena
forms of descriptive studies:
--studies of behavior: facts
   and opinions (studies
   designed to determine
   the current status of
   conditions or attitudes)
--studies of status and
   development (studies
   that trace the genesis
   and change of a group
   of people throughout a
   life cycle
--methodological studies
   (studies that deal with the
   development and
   validation of new tools
   and measuring
   instruments
III.  The Questionnaire
     Survey
     steps:
     A. Selection of
          Questions and
          Instructions
questionnaires: surveys in which individuals are asked to respond to items they have read
          --ways to develop
            survey items:
            use standard
            forms in resource
            works; select
            measurement
            forms that have
            been used in
            research pieces
            you have read;
            develop your
            own measures
         --criteria for
            questions: to
            what extent might
            a question
            influence
            respondents to
            show themselves
            in a good light?
            to what extent
            might a question
            influence
            respondents to
            be unduly helpful
            by attempting to
            anticipate what
            researchers
            want to hear or
            find out? to what
            extent might a
            question ask for
            information
            about
            respondents
            that they are not
            certain, and
            perhaps not
            likely, to know
            about
            themselves?
            --choices in
               question
               forms
            --problems in
              question
              design:  
              double-barreled
              questions,
              loaded
              language,
              improper
              grammar,
              incompleteness,
              vagueness,
              ambiguous
              terms, lengthy
              items, complex
              questions,
              averaging or
              reconstructive
              questions,
              leading
              questions,
              abbreviations,
              imprecise
              questions,
              misspelling,
              awkward
              construction,
              items with only
              one logical
              answer,
              presumptive
              questions,
              elevated
              vocabulary,
              imprecise
              agents of
              action
    B.  Formatting
          (order of
          questions,
          grouping of
          items,
          guidelines
          for length
          of question-
          naires)
direct (that ask for obvious reports) or indirect questions (that ask respondents to react in ways that imply information);
specific (that try to focus attention on individual activities) or general questions (that ask respondents questions) vs. statements (declarations to which subjects indicate how much they agree or disagree)
types of questions
open ended questions: questions to which people respond in their own words
closed ended questions: questions to which people respond in fixed categories of answers
paired comparison questions: questions that ask respondents to make a judgment between alternatives taken two at a time
contingency questions: questions asked only of some respondents, determined by their responses to other questions
ranking questions: closed ended questions that ask respondents to rank order a set of options
inventory questions: closed ended questions that ask respondents to list all reactions that apply to them
matrix questions: closed ended questions that ask respondents to use the same categories to supply information
multiple choice questions: closed ended questions that ask respondents to select a category response from a range of possible responses
     C. Determining
          Reliability
          and Validity
validity: the consistency of a measure with a criterion, which reveals the degree to which a measure actually measures what is claimed (to show that a set of questions is valid, one first must show that it is reliable)
reliability: the consistency of a measure with itself
          --additional
            controls:
            1. check
                questions
 

check question: asking the same question twice at different locations in the questionnaire, usually once positively worded and once negatively worded

            2. measures
                of test
                taking
                behavior
social desirability: a measure of the degree to which people attempt to describe themselves in ways that they think are acceptable, desirable, or approved by others
MMPI Lie Scale: a scale to identify respondents who are attempting to avoid being candid and honest in their responses
Infrequency Index: a measure of the inconsistency of response
            3. polarity
                rotation
polarity rotation: avoiding phrasing all items positively and avoiding placing all positive adjectives on the same side of the measurement items
--response set: a tendency
   for subjects to follow
   predictable patterns of
   responding to test items
     C. Sampling Subjects
     D. Administering the
          Questionnaire
     E. Analyzing and
          Interpreting Results
--survey research reveals
   relationships, but not
   causal relationships
   (experiments or long-
   term historical studies
   are required to make
   those sorts of claims)
IV. The Interview Survey
      --reasons for
         interviews:
        1. helpful when
            permission to
            participate is
            required;
        2. desired to reduce
            refusal rate;
        3. useful when it is
            important to
            record
            information (such
            as a respondent's
            manner and
            nonverbal actions)
            that might be lost
            by using
            questionnaires
    A. Selection of
         Questions
         --types of interviews:
           structured
           interviews
           and unstructured
           interviews
        --recommendations
           for interviewers:
           1. questions
           should be brief
           and to the point;
           2. words and
           phrases should
           be simple and
           in common
           day-to-day use
           among all
           groups in the
           community; 3.
           questions
           should not
           include words
           that have strong
           emotional
           content; 4.
           questions
           must avoid all
           possible bias
           or suggestion
           in favor of or
           against a
           particular point
           of view; 5.
           questions
           should include
           all the important
           alternatives that
           may emerge on
           a given issue;
           6. when the
           individual is
           asked to choose
           between different
           alternatives, this
           choice of
           alternatives must
           be given as early
           in the interview as
           possible.
structured interviews: use of specific lists of questions
unstructured interviews: use of items that permit respondents to indicate their reactions to general issues without guidance from highly detailed questions
strategies for initial questions:
  funnel questions: starting
  with an open-ended
  question and following up
  with increasingly narrow
  questions
  inverted funnel questions:
  starting with a very specific
  question and expanding by
  asking increasingly general
  questions
follow-up question strategies:
  mirror questions: questions
  that repeat previous
  responses to elicit
  additional information
  probing questions: directly
  asking for elaboration and
  explanation
  climate questions: asking
  respondents to explain how
  they feel about the
  interview climate and their
  willingness to answer
  questions
        --a question of ethics:
          confidentiality
          and anonymity
confidentiality: protection of the identity of research participants
anonymity: protection of research participants be separating specific identities from the information given
   B. Training and
        Controlling
        Behavior of
        Interviewers
   C. Determining
        Reliability and
        Validity
   D. Sampling Subjects
   E. Analyzing and
        Interpreting
        Results
        --credibility of
           questionnaire
           and interview
           data:
           data may not
           be trusted if:
           (1) respondents
           may not be
           aware enough
           to give useful
           responses;
           (2) recall is not
           known to be
           accurate
V.  The Focus Group
      Survey
       --purposes: to
         gather preliminary
         information;  to
         help develop
         questionnaire
         items; to
         understand a
         particular
         phenomenon;
     A. The Focus Group
          Method
         1. Assembling the
             Groups
             --not random
               samples, but
               specific
               characteristic
               groups;
               recommended
               size from 6-12;
               more than one
               group typical
focus group: an interview style designed for small groups . . . focus group interviews are either guided or unguided discussions addressing a particular topic to test preliminary ideas or plans of interest of relevance to the group and the researcher
         2.   Preparing Study
               Mechanics
telefocus session: focus group sessions completed by teleconferencing methods over the telephone
         3.  Preparing Focus
              Group Session
              Materials and
              Questions
              --use of the
                presession
                questionnaire
                to increase
                willingness to
                speak
         4.  Conducting the
              Session
              questions not
              restricted to the
              moderator's
              guide;
              moderator's
              exercise skill to
              deal with
              different sorts
              of people; hosting
              single or many
              meetings varies
        5.  Analyzing Data
              and Preparing a
              Summary Report
   B.  Focus Group
         Advantages
         1.  flexible;
         2.  very helpful in
               pilot studies;
         3.  promote frank
              and honest
              sharing of ideas
    C.  Focus Group
          Disadvantages
          1. cannot
               generalize
               results from one
               focus group to
               an entire
               population;
          2.  focus groups are
               only as good as
               the ability of the
               moderator
          3.  as a result of
               group
               discussion, many
               sets of reasons
               and comments
               tend to become
               more and more
               extreme
VI.  Network Analysis
network analysis: a set of research procedures for identifying structures in social systems based on the relations among the system's components rather than the attributes of individual cases
       A. Traditional
            Network
            Analysis
traditional network analysis: a method that obtains individuals' reports of their communication activities with others for the purpose of observing and describing the flow of information in a particular organizational system
            1. The Traditional
                 Approach of
                 Network
                 Analysis
                 purposes:
                
1. to construct
                     a map of
                     the inter-
                     action among
                     people in an
                     organization
                 2. to diagnose
                     problems
                     in communi-
                     cation flow                  
                 3. to identify the
                     roles played by
                     different group
                     members
grapevines: informal flows of communication
            2.   The Method of
                 Network Analysis
                 --general steps:
                   1. selecting
                       communica-
                       tion variables
                       of interest;
                   2. securing
                       individual
                       reports from
                       all members
                       of the group
                   3. constructing
                        network
                        maps and
                        tracking
                        information
                        flow

       

patterns:
reciprocal pattern: one in which individuals share nearly an identical network pattern among themselves
liaison: a person who links people of different networks together
isolate: an individual who is not actively involved in any established network
                   4. interpreting
                        results by
                        comparing the
                        network to a
                        desired
                        standard
       B.  Network Analysis
             Beyond the
             Traditional
             --network analysis of
                social units or
                nodes can be
                people,
                departments,
                organizations, or
                message elements,
                such as words
            
            1. Semantic networks semantic networks: examinations of the relationships among words in a message
            2. Networks among
                Groups
networks among groups: examinations of relationships among groups in an organization
            3. Networks among
                Nations
networks among groups: examinations of relationships among nations
       C.  Limitations of the
             Approach
             1.   emphasizes
                  structural
                  information
                  rather than
                  content
                  information;
              2. does not
                  explain the
                  reasons for the
                  relationships;