Method used to describe and interpret the characteristics of a recorded or visual message

4 categories

1. Rhetorical Criticism

2. Content analysis

3. Interaction analysis

4. Performance studies

verbatim recordings of actual communication

Ex. Recording of court documents

messages produced by communications themselves

ex- written artifacts (letters)

Can be scripted/unscripted

Public/private

Systematic method for describing analyzing interpreting and evaluating the persuasive force of messages embedded within texts

5 functions:

Purposes of messages

Contexts of messages

Social criticism

theory building

pedagogical (teaches)

investigate spoken accounts of personal experiences to understand more fully what happened in the past.

-Interviews with participants of past events

examine public/private texts of prominent, influence or otherwise remarkable individuals

- how msg used by people helped them accomplish what they did

Examines persuasive strategies used to influence the historical development of specific campaigns and causes

-rise and fall of religions/cults/political parties/civil rights activism

Evaluates whether the most appropriate and effective means (as stated in Aristotle's Rhetoric) were used to create the rhetorical texts(s) intended to influence a particular audience.

goal- discover whether speaker makes best choices from inventory to get a favorable decision from a group in a specific situation

1. Invention- develipment using ethos, pathos, pathos and logos.

2. Disposition (organization)

3. Elocution (style)

4. Delivery (nonverbal presentation)

5. Memory

1. Forensic- past concern- legality and justice

2. Epideictic- present cerimonial

3. Deliberative- future

New genres= apologias, self defense, etc.

5 elements of a dramatic events

(dramatic criticism)

1. Act- particular message

2. Purpose- reason

3. agent- person who communicates the message

4. Agency- Medium

5. Scene- context

Based on the work of Ernest Bromann- examines the common images used to portray narrative elements of situations described in a text

Mythic stories present in communication that invovle characters with which ppl identity

Categories

1. Fanatsy Themes

2. Fantasy Types

3. Rhetorical visions

4. Rhetorical Communities

analyzes how conceptions of gender are produced and maintained in persuasive messages

-people's conceptions of the characterisitics of men and women are shaped by rhetoric

Systematic and replicable examination of symbols of communication- which have been assigned numeric values using statistical methods in order to analyze the communciation

Goals: describe and count the characteristics of messages embedded in texts

1. Unobtrusive technique- text already exists

2. Unstructured material- able to categorize

3. Context- study data as it appears in context

4. Handles massive amounts of data

(aka character units)

involve some physical or temporal unit  referred to or alluded to within content... can be used to measure the meaning attached to a particular person, event or issue

1. Selecting texts

2. Determining the unit of analysis

3. Developing content categories

4. Coding units into categories

Analyzing the data

the process of dialogic engagement with one's own and others' aesthetic communication through the means of performance

Researchers interpret texts by carefully examining their communication nuances and then recreating them through performance

Framework

(the science of SM)

To measure results in social media:

Strategy>Management>Strategy

Strategy= goals, everything steams from goals, Not just "lofty things your come up with" The nucleus

Focus on Interactions and engagement

(sceicne of SM)

"Have relations with out publics" Get back to the basics of PR

Judging # of likes, comments and impressions in FB and Twitter is not the best way to go about it.

Measures with litte significance w/o considing the realtionship

Use a coding system tailored to your needs- "What metrics do I care about?"

Berelson's Definition:

A research technique for the objective, systematic and quantitative description of manifest content of communication.

(manifest content= not reading between the lines)

Content analysis

(krippendorff's def)

Krippendorff emphasizes "inference"

is not complete with just quantitative information, must dig deeper and look at inference

Look at qualitative calues

Analyze large data sets (samples)

Data readily available

Conducted quickly/cheaply

Unobtrusive method of analysis (observing recorded comm, don't have to start from scratch)

Accounts for content of items

Generalizable

No IRB

Communication has to be recorded

Rigid categories may overlook insight

Difficulty implementing (intercoder reliability)

·      Set up different archiving systems

·      Create Definitions

o   Define what should be coded

§  Specifics-want US Navy not Old Navy

o   Define what should be excluded

o   Define Categories

·      Train Coders

·      Asses intercoder reliability

Kaid content analysis steps

**know this**

·      1. Formulate Research Questions & hypotheses

·      2. Select sample

·      3. Define categories

·      4. Outline coding process, train coders

·      5. Implement coding process

·      6. Determine Reliability and validity

·      7. Analyze results

1. Formulate research questions and hypotheses

(Kaid step)

·      Pose a question  or a statement

·      Research Question- when trying to prove something and show results. Pose a question.

o   What impact did media access on the coverage of the crisis?

§  Regression model accented for 20% of variance: direct access to personnel on-scene and synchronous mediated access to on-scene personnel predicted a positive tone in media coverage

·      Hypothesis- Give when you know what the answer might be- educated guess- pose a directional statement

o   The challenger would use more attacks than the incumbent on the campaign blog

§  Supported X2: 70.6% of the attack posts came Kerry

§  Bush’s posts- 72.2% contained an Attack (pol. Stmt.)

§  Kerry’s posts- 89.5% contained an attack (pol. Stmt.)

2. Select sample

(kaid step)

·      Determine universe

o   Define what it is you want to sample- specific

·      Pull sample systematically

o   Representative of universe

§  systematically to areas that you have defined

o   Big enough sample to tell story

§  wide variety including all categories

o   Types: stratified sample, constructed week (news coverage from mon-wed)

o   Is there a sampling rate?

·      Determine “unit” of analysis

o   User posted items or Owner posted items

o   Is it a link, blog post, photo, wall post, comment, retweet?

o   Number of likes, comments etc.

o   1 comment=1 unit

o   Define what your unit is to determine what you are coding

·      Write an “inclusion” statement to guide item selection

·      Create a log to keep track of items in sample

·      What is said

·      how it is said

·      Mutually exclusive

·      Exhaustive

·      Create code book & code sheet (see examples in course pack)

4. Outline coding process, train coders

(kaid step)

·      Training session- go over code book

o   Fill out coding and talk about how to fill out a code sheet

o   Make sure human coders are consistent

·      Explain relevant examples from sample

·      Code items, reconcile

5. Implement coding process 

(kaid step)

·      Assign items to coders from log

·      Monitor intercoder reliability (ICR)

6. Determine Reliability and Validity

(kaid Step)

·      5-7% of items “double coded” for ICR

           Double code those to check for reliability

·      Holsti’s formula is percentage of agreement

·      Other formulas: Scott’s Pi, Cohen’s Kappa

·      ICR=.85+ (85% agreement of better)

•  If coder is not meeting 85% threshold, must train coder to be better or fire them
•  ICR= # correctly coded/total # of items (Holsti)

7. Analyze results

(kaid step)

·      Compare historical results to results found

·      Chi square tells if relationship among two nominal categories is significant (probability < or = .05)

o   have 95% confidence can be repeat in another group- accept 5% level of error

·      Correlations determine if items are related at significance rate (P < or = .05)

·      Computer-assisted CA optional

Inter Coder Reliability

ICR

Reliability of findings between coders of a data set.

.85 or better (85% agreement)

Hiosti ICR= # correctly coded/total # of items

Historical Criticism

Neo-Aristotelian criticism

genre criticism

Dramatic criticism

Metaphoric criticism

Narrative Criticism

Fantasy theme analysis

Feminist criticism

1 Physical Units

2Syntactical

3Referential

4Propositional

5thematic

(2-5)= meaning units b/c they involve symbolic meaning and are less standardized