Research for Effective PR Research in PR Public relations is a management function when research is used as part of management decision-making. Process begins & ends with research: RACE: &endash; Research, &endash; Action (planning), &endash; Communication, &endash; Evaluation (show results).   Systems Theory & Research Adaptive subsystem uses research to bring information & feedback to the system. &endash; Secondary research &endash; Primary research Role in policy formulation.   Problem solving Need best available evidence for problem solving (or seizing opportunity). &endash; I.D. problem. &endash; Set goals--situation at end of campaign. &endash; Set objectives--intended outcomes of communication with each public. &endash; Communication and action to reach objectives. &endash; Evaluate how well objectives met.   Roles of research Research is done to define the problem; The basis of planning programs. Want valid, reliable understanding & answers. Issue forecasting. Learning about publics Prioritize & interpret behavior. o Planning media use. o Considering possible outcomes. Once problem defined, then can set goals & objectives. Goals are planned long-term outcomes Objectives are realistic, specific, measurable. Research to monitor program for &endash; Accountability, &endash; Adjustments, &endash; Program changes, &endash; Whether program is reaching target publics. &endash; Feedback to keep program on course. Research to measure &endash; Program impact for success or failure, or &endash; Program's effectiveness in reaching goals and objectives.   Problem detection Informally: Something not quite right. &endash; Discrepancy between observed & desired states. &endash; Opportunistic. Feedback from personal contacts, key informants, community groups, advisory groups, call-in lines, mail, field reports, media. Escalate rigor of surveillance: Qualitative to quantitative TOP Problem exploration Probe environment to confirm the presence of a problem and to expand understanding of problem. Organizational goals are criteria for judging if and when problem exists. Use formal research to explore, confirm and describe the problem.   Formal methods Scientific method Must be: Measurable (no morality, values) Observable Objective (no bias) Replicable Results must be accurately transmissible Essence is comparison   Primary research Quantitative research--aggregates, patterns &endash; experiment &endash; survey &endash; content analysis &endash; panel study (same group over time) &endash; cohort study (random within same group) Qualitative research &endash; historical research &endash; field research--participant observer; case studies; behavior observation; interviews &endash; focus group &endash; legal research   Identifying problem To avoid bias follow rules of research to systematically and objectively gather information: &endash; to increase our knowledge of situation, &endash; to define the problem. &endash; to learn how to best communicate this information to key publics, and &endash; to present evidence persuasively.   Defining the problem Problem statement very important. &endash; Must not suggest one cause at the exclusion of possible others. &endash; Must not suggest one course of action at the exclusion of possible others. Determines the focus & direction of planning.   Problem statement Written in specific and measurable terms. Summarizes what research says about &endash; the source of the problem, &endash; where it is a problem, &endash; when, who it involves or affects, &endash; how it involves or affects them, and &endash; why it concerns your organization and publics. Describes the current situation.   Identifying publics Select & assign priorities to target publics. &endash; Geographic boundaries; &endash; Demographics; &endash; Psychographics; &endash; Opinion leaders; &endash; Communication behavior.   Program goals and objectives Write program's goals: &endash; General. &endash; Ultimate outcome of the program. &endash; Direction of the program. Write program objectives for each public. &endash; Steps to achieving goals. &endash; Specific results to be achieved, &endash; By a specified date. &endash; State in measurable terms how much change is to occur or level to be maintained. Each objective usually requires a strategy. TOP Formative & Summative Research Documentation for accountability is summative research &endash; Assess the effectiveness of program or its parts. Research to detect problems and determine whether adjustments are needed is formative research.   Formative research Content analysis &endash; Comparing content with objectives to assure Messages address issues related to objectives; Messages are consistent across target publics. Estimating readability Tracking output &endash; Distribution vs. placement Enough of right message is being used.   Summative Research To establish a record of time, expenditures, activities and efforts. Evidence for accountability. To document cause and effect (CAUTION). &endash; Evidence to test theories about causal relationships. &endash; Build your own and the profession's knowledge-base. Research in Evaluation Focus on outcomes: How effective was program in achieving its objectives. Changes in attitude or opinion Public opinion surveys Changes in behavior Whether message caused the effect (opinion or behavior change) observed. Experiments. Measuring campaign results When objectives call for behavioral outcomes, use only measures of those specific behaviors.   Reliability & validity Reliability -- Consistency of results &endash; Requires replication. &endash; May be reliable but not valid. Validity--Degree to which instrument or design measures what it's supposed to. Internal validity Degree to which you can conclude X --> Y Most important for experiments. External validity Generalizability--degree to which you can generalize results to population. &endash; Requires randomization. o Equivalent groups; o Equalizes intervening variables; o Every person in population has equal chance of getting picked. &endash; Most important for surveys.   Experiments Essence of scientific method is comparison. &endash; Experimental group compared with control group. Thing being studied must be measurable & observable. ® Independent variable The stimulus (message); the cause of any resulting effect: &endash; Manipulated (exposure to message: present or not). &endash; "X" ® Dependent variable The effect (attitude adjustment, behavior): &endash; Measured to determine if X --> Y. &endash; "Y" ® Intervening variables Other causes: History -- Historical events during course of experiment; Maturation -- People change, mature, grow tired, sleepy, bored, hungry; Testing--Hawthorne effect, test sensitivity; Regression to mean -- changes occur when subjects start out in extreme positions: May be attributed erroneously to X. Interferes with internal validity. Control with randomization & control group. TOP Experimental design Quasi-experiment (or one-shot case study) X Y ----------------------->   Experiment X Y [experimental or (e)] ----------------- .......... Y [control or (c)]   Add randomization for internal validity. X Y (e) R --------------- > ...... Y (c)   Three Experimental designs 1. Static group comparison (post-test only) Y(e) &endash; Y (c) = effect of X X Y (e) R -------------------- > Y (c) 2. Classical experiment (pre-test, post-test) &endash; Pretest interferes with internal validity (test sensitization). Y(b) X Y(a) (e) R -------------------------> Y(b) Y(a) (c)   Solomon 4-Group Design Compare all 4 groups Y(b) X Y(a) (e) R -------------------------> Y(b) Y(a) (c) X Y(a) (e) R -------------------------> Y(a) (c)     Public Opinion Surveys Describe population too large to observe. Measure independent variable (rather than manipulate). Measure knowledge, attitudes, opinions of target public (and control group public). Correlational: Changes in attitude, etc., not scientific proof of causation. Tool in experimental design. Non-experimental use in research &endash; Descriptive (mostly), atheoretical (male, female) &endash; Exploratory (how many interested in issue; voting intention; ) &endash; Explanatory (voting intention tied to characteristics) Measures &endash; Relationships between variables (more TV violence correlated with increased aggression) &endash; Differences (male, female) &endash; Changes in attitudes Before-campaign surveys provide data for objectives specifying changes in public opinion. After-campaign surveys provide data to compare opinion change in test & control publics. Questions in pre- and post-test questionnaires are worded exactly the same. Types of surveys 1. personal interviews &endash; control &endash; good response rate &endash; expensive 2. telephone &endash; cheaper &endash; less intrusive &endash; 80-90% response rate Types of surveys, continued 3. mail &endash; economical &endash; Low response rate &endash; best for complex, long questions Sampling Respondents' characteristics, attitudes and orientations reflect larger population. 1. Non-probability sampling &endash; quota &endash; man on street &endash; convenience &endash; judgment (snowball) 2. Probability sampling &endash; Includes o Simple random sampling o Systematic sample--every nth o Stratefied sample--population elements &endash; If less than 100, use all &endash; Degree of representativeness is function of sample size (to a point) Three kinds of error &endash; 1) response error (interviewer, questionnaire) &endash; 2) non-response error (incalculable) &endash; 3) sampling error (calculable--tied to sample size: larger sample, less error) Sampling error example ® ± 5 percentage points ......45% - 55% ...... / \ / \ 40% - 50% 50% - 60% 45-50 with 5% undecided ..... 45% - 50% / \ / \ 40% - 50% 45% - 55% ....(+5%) (+5%) Measures Evaluative questions: attitudes and beliefs &endash; 1. multiple choice &endash; 2 agree-disagree scale (Likert) &endash; 3. Semantic differential (opposite adjectives) &endash; 4. Rankings Facts (demographics) Information questions &endash; 1. how much respondents know o multiple choice comprehension o open-ended &endash; 2. self-perception questions o opinions and behavior o problem recognition o level of involvement o constraint recognition o information processing Back to top Back to lectures index