To enhance the explanatory richness of the data, the researchers would most likely use

Which experiment occurs when the researcher cannot fully control assignment of subjects to condition?

Tenure has come under attack recently; but there are at least 2 reasons why universities award tenure

economics and academic freedom

Which interview format would be best for the following:

Suppose you are interviewing caregivers of older adult family members with dementia. Which interview format would be best to use when you have the following to consider: 1) when the researcher does not know all the possible answers to a questions; 2) when the range of possible answers is so large that the question would become unwieldy in multiple choice format; 3) when the researcher wants to avoid suggesting answers to the respondent; 4)when the researcher wants the answers in the respondent’s own words

use the following information:

1)      30 students watch a videotape about the Challenger space shuttle disaster

2)      After watching the videotape, all 30 students complete a questionnaire

Which experimental designs BEST describes statements I and 2?

A factorial experimental design can feature or be characterized by all of the following EXCEPT:

full experiment

multiple independent variables

a quasi-experiment

single independent varibale

A 4 x 2 x 2 factorial design is employed to examine the effects of AGE and GENDER on communicator style

1) Identify the number of independent variables 

2) Identify the total number of levels of the independent variables

3) Identify the total number of factors 

1. 3 independent variables (age, gender, and another communicator style)

2. 8 levels

3. 3 factors

survey researchers often use CORRELATIONAL research designs, which are BEST characterized as establishing

While prescreening possible research participants, your research team learns that 20% of the participant pool has completed 5 questionnaires already, and this situation primarily illustrates the

  Intercoder reliability is the most common method for calculating _________ among the observations of more than 1 independent coder

The hypothesis, "Men and women self-disclose differently" is an example of a...

_____________conducted while a program or product is in the process of being developed to identify ways to refine it, and _______is conducted after a program or product is completed to learn its overall effectiveness, usually to determine whether to continue or discontinue it.

Formative

Summative

    Questionnaires and interviews are structured in a variety of ways, depending on the type and arrangement of questions used. The ____________ begins with broad, open questions followed by narrower, closed questions

All members of the population have an equal chance of selection.

1) Construction of a list of all members of the population.

2) Use a method such as dice or coins to select randomly from the list the # of members required for the sample.

You must set up some process or procedure that assures that the different units in your population have equal probabilities of being chosen

Simple

Systematic

Stratified

Cluster

3SC

·         Relies on arranging the target population according to some ordering scheme and then selecting elements at regular intervals through the ordered list

Stratified Sampling

Population embraces a # of distinct categories, categorizes a population along a characteristic (stratification variable)

                  Selecting respondents from certain areas or time periods only. “two-stage sampling” the first stage a sample of areas is chosen; the second stage a sample of respondents within those areas is selected.

doesn’t involve random selection. CANNOT be used to infer from the sample to the general population.

Convenience

Volunteer

Quota

Purpose

Network (Snowball)

(think CV-QP, remember N for NONrandom)

Involves the sample being drawn from that part of the population which is close to hand. Sample population selected bc it is available and convenient.

Another way of saying self selected sampling, happens when volunteer becomes party of the study

Method for selecting survey participants; first segmented into mutually exclusive subgroups, judgment is used to select the subjects

Researcher chooses the sample based on who they think would be appropriate for the study; used when there is a limited # of people that have expertise in area being researched,

Subjects recruit more subjects to the sample; first respondent refers a friend, then he refers a friend etc.

Original documents, journals, theses, dissertations, scholarly texts, conference papers

First reporting of a research study by the persons who actually conducted the study.

Data has been taken from somewhere and has been interpreted

Report of a research study by someone other than the person who conducted the study or a later report by the person who conducted the study that cites or used the primary report that has already appeared elsewhere.

Examples: Textbooks, Books, Magazines, TV, and other (mass media).

reviewers are NOT told the name or institutional affiliation of the persons who submitted the manuscript.

Demonstrates the highest degree of control because the I.V.  is manipulated by the researcher (as opposed to being observed)

Research participants are randomly assigned to create two or more conditions

Either manipulate or observe the I.V. and may have one or more conditions.

When multiple conditions, research participants are not randomly assigned to them so full equivalence is not achieved. Quasi-equivalent conditions are created using pretests. 

In one condition cases, quasi-experiments use multiple pre and posttests as baseline. 

Least amount of control.

Manipulate or observe the I.V. and have one or more conditions.

LIKE quasi-experiments (with multiple conditions), research participants are not randomly assigned to them.

UNLIKE quai-experiments, initial differences between conditions are NOT assessed by using pretests which means conditions must be assumed nonequivalent.

With one condition, preexperiments use a single pretest.

type or category, name. (e.g., male and female, yes and no, or numbers- PHONE NUMBERS OR LICENSE PLATE NUMBERS). Categories in nominal must be mutually exclusive, equivalent, exhaustive.

RANK, “greater than” or “less than”, how much.

Examples: educational attainment;

0 = less than H.S.

1 = some H.S.

2 = H.S. degree

3. some college

4. college degree

5. post college

(explanatory) thought to influence changes in another variable, also called a predictor variable.

changed by another variable; also called criterion or outcome variable, b/c changes in it are dependent on changes in other variables

List all the questions an interviewer is supposed to ask. Interviewers must follow that schedule consistently so that interviews conducted by different interviewers and with different respondents are all done in the same way.

Interviewers ask a set of basic questions on the interview schedule, but they are free to ask probing follow-up questions as well, usually to gather specific details or more complete answers

interviewers are provided with a list of topics but have maximum freedom to decide the focus, phrasing, and order of questions.

(representative values) describe the center point of distribution of quantative data.

Mode- simplest, indicates score or value in a distribution occurs most frequently.

Median- divides a distribution of quantative data exactly in half.

Mean- average, computed by adding all the scores in a distribution of interval/ratio data and dividing by the total number of scores.

Measures of variability, report how much scores vary from each other or how far they are spread around the center point of a data set and across that distribution.

Applied to ORDINAL, INTERVAL, AND RATIO DATA. Range (span) simplest of dispersion, reports the distance between the highest and lowest scores in a distribution. Subtract LOWEST from HIGHEST

average distance of the scores in a distribution of interval/ratio data from the mean in squared unites

measure of dispersion that explains how much scores in a set of interval/ratio data vary from the mean, expressed in the original unit of measurement. TAKE THE SQUARE ROOT OF THE VARIANCE.

Threats to how research conducted, participants, researchers

 1) Single Group Threats: History (changes in external environment), Sensitization (initial testing affects later testing), Instrumentation (changes in data methods), Maturation (phyiscal/psycholoical changes in participants), Mortality (participants drop out of study), Selection (methods for obtaining participants)

2) Multigroup Threats: Selection Bias or Threat - 2 groups were not comparable before the study (can have selection-history, selection-sensitization, etc.)

3) Social Interaction Threats: social pressures in research context, not caused by the treatment, b/c of groups or key people in research

Imitation Treatment (1 group imitates the other), Compensatory Rivalry (1 group developes a competitive attitude), Resentful Demoralization (opposite of Compensatory)

One-Group Posttest-Only Design 

(FOR PREEXPERIMENT DESIGNS)

One-Group Pretest-Posttest Design

(FOR PREEXPERIMENT DESIGNS)

Posttest-Only Nonequivalent Groups Design (Static Group Comparison)

(FOR PREEXPERIMENT DESIGNS)

Single-Group Interrupted Time Series Design (Time Series Design)

(QUASI-EXPERIMENT DESIGN)

Pretest-Posttest Quasi-Equivalent Groups Design

(FOR QUASI-EXPERIMENT DESIGNS)

Interrupted Time Series Quasi-Equivalent groups Design (Multiple Time Series Design)

(FOR QUASI-EXPERIMENT DESIGNS)

Combines the previous 2 quasi-experimental designs by nonrandomly assigning participants to a treatment or control group and measuring them on a series of pretests and posttests.

Pretest-Posttest Equivalent Groups Design

(FOR FULL EXPERIMENT DESIGNS)

Posttest-Only Equivalent Groups Design

(FOR FULL EXPERIMENT DESIGNS)

same as pretest, posttest equivalent group except that a pretest is not used; participants are randomly assigned to a treatment or control group and

            given a posttest.

Solomon Four Group Design

(FOR FULL EXPERIMENT DESIGNS)

literally combines the pretest-posttest equivalent groups and the posttest-only equivalent groups 

Factorial Designs are used when there is more than one I.V. being studied. 

add all the scores and divide by the total number of scores.

square root of variance.

take the mean, each individual score – mean, than square each score- mean. Add upp the sum of squares and divide by the number of scores

Correlation indicated whether 2 variables are related and if so, how much. By itself correlation is susceptible to alternate causality arguments. Sequencing of events in time to infer causation. Correlational designs use a questionnaire or interview to assess all the variables of interest at one point in time and then analyze the relationships among them.

Statistical relationship between variables; type and strength of a relationship between variables

Correlation is one of the 3 criteria used to determine causation.

Causation can sometime be inferred from such a study, survey research established noncausal relationships among variables.

IV's produce an outcome of the D.V.

I.V. must precede the D.V. 

I.V. & D.V. must covary (go together)

Changes observed in the D.V. must be the result of changes in the I.V.

List of population from which they will sample.

Lists all the members of the population, although in actual practice the sampling frame is as exhaustive a list of the population as researchers can obtain.

Random-digit dialing solves some of these problems by having a computer generate randomly all possible combinations of telephone #'s in a given exchange.

Percentage of phone calls that result in contact with an English-speaking interviewee.

Percentage of phone calls in which interviewees agree to participate.

Respondents had been surveyed too frequently, not eligible to participate in the survey.

Regardless of how a sample is obtained and the sampling unit of analysis, survey researchers have to be concerned about this rate. The number of usable responses divided by the total number of people sampled. This is crucial for evaluating survey research. There are important differences between those who did and did not respond, the results of the survey may not be valid.

(direct) increases in independent variable are associated with increases in a dependent variable (e.g., the more hours one spends studying before an exam, the higher one’s exam scores will be).

 (inverse) increases in an independent variable are associated with decreases in a dependent variable (e.g., the more hours one spends partying the night before an exam, the lower one’s exam scores will be).

An implicit statement that underlies every hypothesis and predicts that there is no difference between the groups or no relationship between the variables being studied. Underlies 2 tail hypothesis, represents chance occurrence

One-Tail; Directional  Specific and may predict the MORE specific nature of the relationship (more than, positive, negative, pointing in certain way).

Two-Tail; Nondirectional, General and predict relationship between variables without specifying the nature of that relationship (differently, open ended)

The larger the size of a random sample, the more it approximates the true population and the smaller the amount of error. True up to 1600-2000 people.

Quantitative Data

Numerical data (mean, mode, median, range, variance, standard deviation), deals with numbers, data which can be measured, length, height, area, volume, weight, speed, time, temperature, humidity, sound levels, cost, members, ages, etc.

(think QUANtitaive, QUANtity)

symbols, deals with descriptions, data can be observed but not measured, colors, textures, smells, tastes, appearance, beauty

(think QUALitative, QUALity, things described are the quality of something)

Probability level researchers set for rejecting a null hypothesis. Establishes a point at which a researcher is confident enough of a statistical difference or relationship to reject a null hypothesis.

Specific significance level set by a researcher for rejecting a null hypothesis prior to conducting a study and analyzing the data.

Example:  A .05 significance level is required to reject a null hypothesis, using a 95% confidence level. .10 means 90% confident of rejecting null hypothesis.

Consistency, measured numerically, 70-75% if 7/10 times something has been consistent, you accept it. Does not imply validity.

o   NOT measured, it is estimated.

o   Test/retest (ex: you should get the same score on test 1 as you do on test 2) and internal consistency (estimates by groupings in a questionnaire that measure the same concept)

Implies reliability (accuracy). To measure consistency, observe how many times something happens.

o   best available approximation to the truth or falsity of a given inference, proposition or conclusion. Validity is the strength of our conclusions/inferences.

deduction, mechanical, quantitative, formal, objective

 induction, observational, qualitative, informal, subjective

REALITY

Positivists see reality as singular and objective, only 1 reality exists.

Naturalistic see multiple realities

   Deals with the concern of the relationship of the researcher to that which is being researched.

Positivist = independen relationship

Naturalist = interdependent relationship

    Role of values in the research process.

Positivist = value-free and unbiased, keep researchers values out of research

Naturalist = research is value-laden and biased

Concerns the process (think METHODological) of research.

Positivist = deduction, search for cause and effect, relationships between variables, static design, researcher-controlled setting, quantitative methods, context-free generalizations, and goals of explanation, prediction, and control.

Naturalist =  induction, wholistic understanding of patterns of behavior,

emergent design, natural setting, qualitative methods, context-bound findings, and goals of understanding and social change.

Language of research reports.

Positivist = formal and impersonal (3rd person) voice

Naturalist = informal and personal (1st person) voice

4 Measurment Scales

nominal, ordinal, interval, and ratio.

categorize, rank, EQUAL DISTANCE, include ARBITRARY ZERO POINT scale, LIKERT AND LIKERT-TYPE scale identify extent of a person’s beliefs toward object; it asks whether they agree or disagree on 5 point scale.

categorize, rank, equal distances; establish an ABSOLUTE, or TRUE, ZERO POINT. Zero point scale- NO negative values. (e.g., age, weight, height)

Provide respondents with preselected answers from which they choose or call for a precise bit of information. 

Ask respondents to use their own words in answering questions. 

1-tailed or 2-tailed?

When probed, persons enacting unfamiliar lies will become reticent, while persons telling familiar lies will maintain levels of vocal expressiveness.

1 tail or 2 tail?

There will be significant differences in reported leisure time between individuals with cable-channel access and individuals without cable-channel access.

Conceptualization

Planning & Designing Research

Methodologies for Conducting Research

Analyzing & Interpreting Data

Reconceptualization

Self-Reports -People respond about themselves (problem of social desirability bias)

Others' Reports -People report on others

Observations - People report on others based on observations (behavioral acts)

Questionnaires are most frequently used in comm research. Measures the IV and DV on a long scale survey.

Interviews are oral and can also be used to study behavior.

Basic Research doesn't have to be practical. Used to increase our knowledge about communication by testing, refining, and elaborating a theory.

Applied Research must be conducted in a field and done for a particular audience and for solving a real world problem

Identifies the extent of a person's beliefs, attitudes, or feelings toward some object. Asks people the extent to which they agree or disagree with a statement by choosing a category on a 5 point scale ranging from strongly agree to strongly disagree.

Positivist and Naturalistic Paradigms

Positivists are concerned how to apply methods in physical sciences. Science, Scientific Method.

Naturalists are focused on nature of reality and human behavior. Social.

Changes in the independent variable cause the observed change in the dependent variable. Example - researches believe smoking cigs causes cancer.

variables are associated or occur together without one necessarily causing changes in the other. The cause or effect is unclear.

Respondents are asked a straight series of similarly organized questions.

Example: asking people to identify the television shows they typically watch each night of the week

+1SD and -1SD = 68.26% = People labeled as average

+1SD and +2SD = 13.59% = People labeled as above average

-1SD and -2SD = 13.59% = People labeled as below average

+2SD and 3SD = 2.14% = Poeple labeled as high

-2SD and -3SD = 2.14% = People labeled as low

Normal Distribution tells reserachers the relative probability of a score falling in any given area of the curve