methods ch 6,7,9

Behavior coding pretesting

A third person monitors the researcher asking questions of the participant

Cluster sampling

A two-stage sampling process used when the researcher cannot identify a complete listing of the population

Likert-type scales

All of the above; are a common response set for closed questions, assign numerical values to each response choice, are interval-level data, contain a neutral middle response

Interrater reliability for categorical data

All the above: - Is also known as interrater agreement. - Is generally associated with content analysis - Is also known as intercoder reliability - Can be calculated with Cohen's kappa or Scott's Pi

A good survey item is one that is

Both a and b; reliable and valid, straightforward

Sampling error

Both b and c: based on informational and always present

Purposive sample

Choosing individuals to participate in research because they are considered "typical" of the critical characteristics important to the research project.

Convenience sample

Choosing individuals who are convenient and easily available as research participants

Measuring the core concept that was intended to be measured and not something else

Construct Validity

The degree to which the measuring device represents the full range of characteristics associated with the construct of interest

Content validity

Achieved when one measurement can be linked to some other external measure

Criterion-related validity

Researchers often include stimulus statements with questionnaires or surveys the purpose of the stimulus statement is to

Direct participants attention to the type of questions they will be answering and to provide general instructions for answering the questions

Employment status

Discrete data

Sex

Discrete data

Stratified random sampling

Divide the population according to Homogeneous groups; randomly select participants from each group

Systematic sampling

Divide the population by the desired sample size to establish that every person should be selected; select random number to establish where in the list to begin selection

Convenience samples are never an appropriate choice for identifying research participants

F

Generalizability is the extent to which conclusions developed from data collected from a population can be extended to the sample

F

It is okay to include some items in a questionnaire for which participants might not have adequate knowledge to respond. In such a case, participants will know to ignore these items

F

Regardless of the researcher's level of experience, it is always wiser to develop a new and unique questionnaire or scale rather than modifying or adapting existing questionnaires or scales.

F

Sample size is the same as the number of people you ask or selected to participate.

F

Surveys on the same topic written to achieve different objectives will be composed of the same questions and response choices, and be subject to the same interpretations.

F

The choices in a closed question response set should be exhaustive and mutually exclusive but do not need to be equivalent

F

Just by looking at the measuring device, we believe it will measure what it is supposed to measure

Face validity

Snowball sample

Finding additional research participants by asking current participants to recommend individuals with similar characteristics

Conventional pretesting

Individuals like the persons who will be selected for the research project complete the survey or questionnaire just as participants will in the research project

The relationship between the concept being measured and the process of measuring it

Internal Validity

The degree to which the multiple items provoke the same response from the participant

Internal reliability

Score on communication apprehension questionnaire

Interval data

Ranking of preferences for political candidates

Ordinal

Ranking of television markets

Ordinal data

Probability sampling means that selection of participants is

Random

Number of times the audience asks a speaker questions.

Ratio data

The measurement is stable, trustworthy, or dependable

Reliability

Closed question

Responses must be exhaustive, mutually exclusive, and equivalent

Quota sampling

Seeking a specified number of participants who fit characteristics of sub-groups important to the research project

Using half of the measuring items for one measurement and the other half for another measurement

Split-half reliability

A sampling frame is the available population in which a researcher is interested and from which research participants are selected

T

A significance level of .05 means that 5 out of 100 findings that appear to be valid will, in fact, is due to chance

T

An alternative to a self-report questionnaire is to ask the same questions in an one-on-one interview

T

Non-Probability sampling is used to generate samples for quantitative research projects

T

Significance levels are never set at levels other than .05

T

The smaller the sample, the greater the proportion of participants needed for the sample

T

The degree of similarity in measurements captured at different points in time

Test-retest reliability

The null hypothesis is rejected even though its true

Type l error

The researcher claims some difference or relationship exists when one does not

Type l error

The null hypothesis is retained even though it is false

Type ll error

The researcher misses claiming a difference or relationship exists or relationship that is real

Type ll error

Volunteer sample

Using individuals who offer to participate in a research study

Measures what you want it to measure and not something else

Validity

Reflects true differences among individual's scores

Validity

Open question

When using this type of question in a written survey or questionnaire, leave more space than you believe respondents will need

Hypothesis testing

all of the above except d; relies on sampling and significance levels, is really decision making about accepting the alternative explanation or retaining the null hypothesis, allows for Type l and ll errors,

The significance level

all of the above: is the same as the probability level, can be reported as p in written research reports, can be reported as the alpha level in written research reports, is set by the researcher prior to computing the statistical test

A semantic differential scale

all the above: asks participants to locate the meaning they ascribe to a stimulus, is a form of continuous level measurement, is anchored by two opposite or bipolar adjectives, does not include descriptors for the intermediate positions

measurement allows researchers to make comparisons among

all the above: individuals in the research project, multiple studies using the same measurement device, variables in one study

A threat to validity or reliability is

any data-related problem that could lead you to draw a false conclusion from the data

Maximum variation sampling is

based on informational redundancy; a researcher seeks participants until the data received are the same as previously collected data.

continuous level data

both a and b: reflect differing degrees, amounts, or frequencies-range from a minimum to a maximum quantity

Unitizing reliability and categorizing reliability are used with

content analysis

A population is

determined by identifying the characteristics the researcher wants participants to have.

Simple random sampling

every person has an equal chance of being selected for the sample; individuals are selected one at a time

measurement is

everything the researcher does to arrive at the numerical estimate

Easiest type of validity to establish

face validity

A Likert-type scale

has a response set that must be balanced at the ends of the continuum

Cognitive pretesting

helps uncover questions the can be interpreted in multiple ways

A reliability coefficient indicates the degree to which this concept occurred

idk

A semantic differential scale

is based on two bipolar adjectives

Internal reliability is the degree to which

multiple items are consistent in measuring a construct or subconstruct on a questionnaire or survey.

Likert-type scale

often uses the responses of strongly agree, agree, undecided, disagree, and strongly disagree

a variable measured with discrete data means that the data identify participants as belonging to

one of at least two categories of the same variable

An open question is

provides data from the respondent's point of view; can create responses that are not comparable

Semantic differential scale

responses indicate degree to which they agree or disagree with one or two anchors

A social desirability response is

the answer the participant believes the interviewer or researcher is seeking.

A significance level is

the level of error the researcher is willing to accept for each statistical test.

Closed questions are used most effectively when

the potential responses are known in advance

Expert panel pretesting

those familiar with research methodology or the content of the survey or questionnaire point out potential problems

Recall cues are used

to direct participants' attention to the issue in which the researcher is interested