research methods 390
units of analysis - groups
- - Sometimes communication dynamics change when individuals are placed together in groups to carry out a goal - Kuhn and Poole (2000) focused on 11 groups in 2 major corporations as the unit of analysis in their examination of the effects of group conflict management styles on decision making - Found that groups that were more integrated in conflict styles and addressed conflict openly were more effective in decision making versus groups that engaged in conflict avoidance or confrontational style
measurement procedures and rules
- 2 important considerations in all measurement are the procedures used in observation and the rules used in assignment of number - Ex: want to know how many cars passed thru a certain intersection on any given day, might ask people who we know are often near that intersection, and how many cars they think
staple scale
- An offshoot of the semantic differential scale - Designed to measure individual's attitudes toward a topic by providing participants with a set of adjectives and then having them rank those adjusts from -5(words are not an accurate portrayal of something) to +5 (words are an accurate portrayal of something) - Ex: creating a staple scale for anxiety
variable levels
- Another integral part of research design process involves selecting variables to examine that will have answered the research question or hypothesis - Each variable selected for study will be directly associated with identifying particular statistical method appropriate for answer the research question - Variables can possess qualitative or quantitative characteristic - Variables with qualitative characteristics = involve assigning items to groups or categories to describe population - Variables with quantitative variable = can be easily identified bc they represent different quantity or amount - A research variable classified into 1 of 4 levels: nominal, ordinal, interval, ratio
generalized attitude measure
- Another scale has the ability to measure different things every time it is employed is the Generalized Attitude Measure - This scale measures a range of different attitudes depending on what a researcher needs to analyze (similar to Generalized Belief Scale) - Attitude: defined as a predisposition to respond to people, ideas, objects in an evaluative way; define as a predisposition bc we are defining an attitude as an individual's tendency to do
semantic diff scale
- Asks respondents to rate their opinions on a linear scale between two endpoints that have opposite meaning - Between these two oppositely worded adjectives exists a series of steps - Most common is 7 - Ordinal
likert scale
- Attitude-scaling technique, basic premise was simplistic but revolutionized social scientific research - Started off presenting respondents with a declarative statement of some kind - Ex: cold pizza is a good breakfast o Offered a range of possible choices: strongly disagree, disagree, neither agree nor disagree, agree, strongly agree o 5 choices and pick one category and have weighted distances o Ordinal bc researchers cannot assume that respondents perceive the difference between different levels evenly - Latent variable or hypothetical variable: is a variable that a researcher cannot directly observe, but that can be inferred from other variables that are observed and measured directly
benet- simon scale
- Child's adeptness at completing carious age-appropriate tasks - First known intelligence test
statistical differences
- Difference test only tell us half the story - Two curves can be statistically different but the difference between female and male average score could be a matter of 5 points - Researchers also interested in the size of the difference (effect size) not only the fact that a difference exists
dyads example
- Ex: Barbato, Graham, and Perse (2003) asked 202 parent-child pairs to complete questionnaires assessing perception of family communication climate and interpersonal communication motives for interacting with one another - Results of study indicated that the communication climate had significant influence on the motive selected for communicating in the dyad
example of unit of analysis - individuals
- Ex: study of sports fans, selected the individual as the unit analysis; data collected to identify an individual's level of verbal aggressiveness (attacking the self-concepts of people instead of the position they espouse; e.g., "only an idiot would think that" or "what kind of moron are you?") and its relationship ship to perceptions of the appropriate communication at sporting events - Verbal aggressiveness and perceptions of appropriate behavior use the individuals as the unit of analysis
contextual CA
- Experience oppressions in one of the communicative contexts - interposal, meeting, group public - Ex: someone may experience CA interacting with someone on a one-one-one basis, but have no problem standing up to an audience of 1,000 people and giving a speech
Generalizations and Inferences
- Identifying the unit of analysis = essential to maintain focus during research project and to enable researchers to generalize result to similar units - Important to resist temptation to make inferences about individual based on their group membership - The concept of ecological fallacy is similar to stereotyping and occurs when researchers assume that bc a participant is a member of a specific group (e.g. culture) he or she possess all the same communication characteristics as the group - Don't assume that one organization is similar to all other bc this would lead to erroneous conclusions - Important to keep in mind that units of analysis have unique characteristics
Units of Analysis - Organizations
- If a researcher is interest in identifying how organizational members interact with one another and the resulting impact on task completion, efficiency, satisfaction, or a host of other variables, the organization become the unit of analysis - Example: Workplace Aggression Tolerance Questionnaire to identify perceptions about the appropriateness of workplace verbal aggression - Core of research focus on appropriateness of behavior in the organizational context
likert scale
- In a Likert Scale - participants presented with a number of statements and then asked to respond to those statements based on a pre-existing scale - 5 step ranging from 1 (strongly disagree) to 5 (strongly agree) - So only 5 possible choices - Each individual item can be added together to achieve a composite score, which can then be used in various statistical analyses - PRPSA example
numbers and things
- In measurements, numbers are assigned to represent "things" - And if our measurement is to be of any value, it is important that the numbers and the things are isomorphic - Isomorphism: means the identity of similarity of form - Question we are asking when concerned with whether our measurement is isomorphic with the thing being measure is "Are our rules for numerical assignment tied to reality?" - Ex: estimate level of anxiety felt by a public speaker, as the speaker is speaking we may rate them on scale from 1 to 5 (low anxiety to high anxiety) based on the distracting mannerisms they evidences in speaking - these rankings would be isomorphic with anxiety if distracting mannerisms are positively related to anxiety, but if, as is actually the case, distracting mannerism are more a product of habit and lack of experience in public speaking than a product of internal state of anxiety then our rating are non-isomorphic
sum of squares
- Information about both the amount and the origins of individual differences - The more individual scores in a dataset differ from one another, the greater the sum of squares will become
mode
- Least frequently cited descriptive statistic - Mode: the value that occurs most frequently in a dataset - "mode is the most"
median
- Median: the middle value in a list of data - To calculate the median, first list the data in order from smallest to largest
range
- Most general measure of variability - Represents how far apart scores are from one another, or the distance between the largest value and the smallest value in the dataset - Range calculated by subtracting the lowest score in the distribution from the highest score
interval variables measure qhat
- Most interval variables studied measure traits, behavior, beliefs and attitudes - First examine what personality and communication traits are then look at various communication traits measure using an interval scale - The interval variable included in the dataset are communication apprehension, ethnocentrism, humor assessment, non-verbal immediacy, sociocommunicative orientation, willingness to communicate, beliefs and attitudes
nominal variables
- Nominal variable data: make up the variable level that is identified by its qualitative characteristics - To categorize nominal level data - simply name each characteristic using categories that are mutually exclusive - Important to remember that a characteristic can only be represented by one category of a variable - No logical order is followed and values cannot be assigned to quantitative values or rank ordered - 3 primary rules for categorizing nominal level data: o Attributes must be mutually exclusive o All categories used to represent data must be equivalent o All categories used must be exhaustive - Descriptive statistics and frequency distributions = primary method of data analysis for nominal level data - Make no sense to try to compute a mean score for sex bc categories are mutually exclusive
operationalization
- Once you have determined the basic concept you are actually planning on measuring you must operationalize your concept - Operationalization: the detailed description of the research operations or procedures necessary to assign units of analysis to the categories of the variable in order to represent conceptual properties - Applies to all forms of measurements, not just working with scales - 1st form = manipulation into categories - 2nd way = estimate the category either by observing existing records or asking people - 3rd = interviews and surveys, asking for verbal or nonverbal reports of phenomena
antecedent variables
- Potential to affect relationship found between independent and depend variables - These variables answer the question "What happened before the data were collected?" and shed light on some differences that could emerge - Ex: impact of conflict (independent variable) on martial satisfaction (dependent variable) - Should take into consideration possible antecedent variable that might impact like if both partners had incompatible personalities before marriage could led them to experience conflict in their relationship different to those compatible
process of measuring a line
- Process of measuring a line - First, define what the phenomenon is that we are attempting to measure - With a line, attempting to measure the length of something - Once we know what the phenomenon is we can determine how we will assign numbers to that phenomenon - With length, many numerical categorizing systems
interval variables
- Quantitative and classified in a logical order that represent equal distances between levels within each category - Example: levels of people's public speaking anxiety, levels of perceived similarity - Enable researcher to rank order items and to compare size of difference between each category
differences od degree
- Quantitative researchers are interested in this - Differences of degree: when 2 groups have differing degrees of a variable they both display - Example: determine whether females and males differed in their levels of verbal aggression; give groups a survey to fill out then look at averages of the 2 groups - Statistical difference tests would allow researcher to determine whether or not the 2 groups have different average scores for verbal aggression - Eyeballing 2 curves may make it look like it's the same but can't use
scalogram or guttman
- Scale that is used to create an interval variable; rarely used - Consist of a set of unidimensional items that respondents rank in order from the least extreme to the most extreme position - Ex: determine how much homework is too much homework for high school students; give them a list of 10 possible time frames ranging from 0 to 10 hours; then have participants answer yes or no next to each frame - The goal of this is that at some point you will hit a cutoff point beyond which people will not go
semantic differential scale
- Second type of research scale commonly used to examine interval level measurements = semantic differential scale - Consists of a series of adjectives that are oppositely worded o Ex: good/bad, happy/sad, eventful/uneventful, positive/negative, etc... - There then exists a continuum of possible choices between the adjectives and the participants are asked to select the number that most represents their perception between the two adjective - May be asked to rate the move She's the Mean using the bipolar adjectives "good" and "bad" with a choice of 1 to 7 between the two words; numbers 1 and 7 indicate very strong feelings, 2 and 6 indicate strong feelings, 3 and 5 indicate fairly weak feeling, and 4 is undecided - When you add up responses from all adjective pairs, you can use this number in a variety of statistical tests
ratio variables
- Similar to interval variables except ratio variables have an absolute zero starting point to represent the absence of the characteristics - Examples: income, exam scored, temperature (when using Kelvin) and speed
neutral relationships
- Sometimes researcher may discover that the variables selected for a study are not related - Neutral relationships: refer to this lack of relationship, or when the change of one variable doesn't correspond to change with another variable - Ex: pizza one consumes and individuals level of public speaking anxiety
mean
- The mean or average, defined as the value that represents an entire group of scores - To calculate the mean, add all of the scores in the category you wish to summarize and then divide the total by the number of scores - Formula: population mean = sum of all scores within a population / number of scores within a population
ethnocentrism
- The view that customs and practices of one's own culture are superior to those of other cultures - Term comes from combo of 2 Greek words: ethnos (nation) and Kenton (center) - When one holds ethnocentric views, as virtually everyone does, an individual's culture is used as the standard by which all other cultures are evaluated - In fact, any deviation from individual's culture is most likely (but not always) seen as negative, which indicates the inferiority of the other culture and the people from the culture - Can be positive - that it gives the people in a culture an identity and helps make them more cohesive, which in turn helps promote positive and effective communication among members of
types of variables
- To answer the research question or hypothesis, the next step involves identifying 2 types of variables chosen as the focal points of the study - Variables can be classified as: independent, dependent, intervening and antecedent
measures of central tendency
- To measure the "center" or "middle" score in a group of data - Can be examined on a frequency distribution are the mean, median and mode
mental measure
- Tools used for the measurement of mental functions attitudes, beliefs, cognitive knowledge, perceived knowledge, and personality/behavioral traits - Any achievement test, personality test, aptitude test, intelligence test, or career choice test qualifies as mental measure - Mental measure known as research scales - Develop measurements for nonphysical attributes - Measure simple mental processes like the time it took people to perform simple mental acts like naming objects in sequence or color
trait ca
- Trait CA has been shown to impact other areas as well - High CA individual tend to have low self-esteem and more prone to depression and perceived as having low intelligence by peers
nominal
- assign numbers to objects where different numbers indicate diff objects - no real meaning other than differentiating between object - ex: male =1 female = 2 - females could be 666 or 777 and males could be 555 or 999 - ex : baseball uniforms
ordinal
- assigning numbers to objects (like nominal) but here these have meaningful order - ex: place finished in a race 1st, 2nd, 3rd and so on - we know how they did, the place they finished and how they did relative to others - indicates order - same in an election - we know exactly how they did in relation to others - number placement or order so we can rank people
what are the 2 aspects of variables?
- attributes - values
variable values
- attributes identify the category for variables - refer to the numerical designation assigned to each variable to allow statistical analysis ex: age, possess a numerical value on their own bc value of variable already exists
variables
- can be defined as any entity that can take on a variety of different values; concepts or construct that varies - ex: age, sex, or level of public speaking anxiety (or fear of public speaking); each of these ______ will take on a different value for diff people at a given time
nominal versus ordinal, interval, and ratio
- ex: baseball uniforms 25-20 = 5 and 35-30 = 5 that doesn't make sense - uniform used only to serve diff players - no order (ordinal) - no meaningful differences (interval) - no ratio or true zero point (ratio) - ex: someone who choose number 0 in baseball not absent of a player just a player with this number
negative relationships
- exist when a decrease in one variable corresponds to an increase in another variable, or vice versa - ex- Example: negative relationship between public speaking anxiety and perceived similarity with one's audience - Increase in one variable (perceived similarity) correspond to decrease in other variable (level of public speaking anxiety)
positive relationship
- exists when an increase in one variable produces an increase in another variable - or a decrease in a variable corresponds to a decrease in another variable ex: found that individual level of public speaking anxiety was positively related to individual's level of conspicuousness - increase in one variable ( level of conspicuousness) corresponds to increase in the other (level of public speaking anxiety)
unit of analysis - individuals
- focus on an individual's temperament, personality or communication traits, or the research might focus on specific communication strategies or behaviors selected
dyads
- focuses on collecting information about two people involved in an interpersonal relationship
ratio variable
- for a variable to be seen as a ratio, the distance between attributes must have real quantitative meaning and an absolute zero - age is the only variable the meets this criterion
interval
- identify distance - equal interbals - temp
factor analysis
- looks for patterns in particpat
dependent variable
- not manipulated - variables recorded or measured but not exposed to any change
interval
- numbers have order (like ordinal) but there are also equal intervals between adjacent - scale of measurements build on - ex: temperature in F diff between 76 and 77 is that same as 17 and 18 - 10lb is twice as much as 5lb (ratios are meaningful 10/5 = 2) and meaningful ratios - zero lbs means no weight or absence of true zero
paper - intrusment
- once you have finished with the first 3 parts of a method section you must explain all of the measurement tools employed in your study - to explain a scale, tell your audience who the creator of the scale was and what its suppose to measure - explain how its formatted, including a discussion of type of scale
belief
- our perception of reality about whether something is true or false
independent variables
- part of the research environment that are manipulated are changed - is what we are studying with respect to how it impacts a dependent vaiable
attitude
- predisposition to respond to people, ideas, or objects in evaluative ways
interval versus ratio
- ratio = meaninful and true zero point - temperature 0 degrees doesn't mean absence of the property of no true zero point and 80 degrees isn't twice as hot as 40 degrees (interval) - weight of objects 5 pounds and another is 5 pounds on a scale it would give you 10 lbs - but if you have a 40 degree day and a 40 degree day you have two cold days and do not end up with an 80 hot day - temperature = interval - weight = ratio
variable attributes
- refer to the specific categories of a variable - ex: biological sex includes the attributes of male or females
differences
- refers to the degree to which one person or a group of people are dissimilar from another person or group of people - differences in kind and differences in degree
cognitive knowledge
- scales that attempt to measure knowledge - t or f - mc exam
nominal
- simple classification - when we assign a person a number - arbitrarily selected and meaningless - ex: gender of students 1 = male 2 = female
concrete variables
- stable or consistent - might include characteristics such as biological sex or sibling order (oldest, middle youngest)
belief vs attitude
- studying beliefs and attitudes is important for communicative researchers - using the word "believe" we imply that for us (not necessarily everyone else) the study o beliefs and attitudes is a true, correct, appropriate set of variable to study - we "believe" in the studying of beliefs and attitudes by communication researchers and also think that studying beliefs and attitudes is good, so we have a positive attitude toward it - belief has to do with our perception of reality - attitude has to do with our evaluation of that reality
interval versus ordinal
- temperature = 1 degree diff is the same as all points of the scale (interval) - place in a race 1st, 2nd and 3rd -- the diff between 1 and 2 not the same as 2 and 3 (ordinal) not equal - ordinal bc order, but not adjacent categories
paper = participants
- the 1st part of the method section should report who the participants were - report basic demographic characteristics (sex and age) and how you recruited them
measurement
- the process of applying numbers to objects according to a set of rules - we apply numbers or give numbers to an object - assigning numbers to something and follow some sort of rules
measurement
- the process of systematic observations and assignment of numbers to phenomena accord to rules - process - - The word "process" is defined as = a set of progressive, interdependent steps; when we measure something we must go through a series of steps - "observe" means we must have the ability to see or perceive that phenomenon; if we cannot see it, there is no way to measure it - Once a researcher has a general idea of what the phenomenon is, we can assign numbers to it based on some set of established rules, way to quantify it
abstract variables
- those that change or differ over time or across situations or context; example: communication satisfaction or self-disclosure
parameter
- value of a population - Ex: administered an intelligence test to all people in the world, we might then add all of their scores together, divide that total by the number of people, and obtain the "mean" or average score for all people on this measure of intelligence - This "mean" would be the parameter of the population - Assume level of measurement employed in obtaining data was at least the interval level
intervening variables
- variables whose presence may impact the relationship between the dependent and independent variables - Intervening variables: are often included in a study to determine whether the effects or a change in the independent variable in turn cause a change in the dependent variable - If you wanted to examine the effect of teacher feedback on student motivation, might be beneficial to consider the variable of self-esteem as a potential intervening variable - Positive feedback form teacher may boost student level of self-esteem which in turn may impact level of motivation
ordinal
- we assign numbers to things in such a way to show relationships amoung thingss, rank-order - direction of difference in some meaningful way - hierarchial level between groups
ratio
- weight - true zero point
paper - prodecures
- when discussing procedures should be as brief as possible while giving all the appropriate and necessary information about what you did - approached, obtained consent, student indicate desire, etc
constructing questions
1.) Start with twice as many items you will need 2.) Every item should reflect the construct 3.) Use concise, clearly worded, unambiguous terms 4.) Construct relatively short items 5.) Pay attention to terminology in the item 6.) Avoid emotionally charged items 7.) Avoid leading items 8.) Avoid loaded items 9.) Avoid double questions 10.) Avoid questions with false premises 11.) Avoid using words "always" and "never" 12.) Avoid double negatives/positives 13.) Avoid hypothetical questions 14.) Avoid ambiguous pronoun reference 15.) Consider recall issues for certain types of item
Pancultural
: few things are found to be ___ (the same across all cultures), but this view that an individual's own the same across all cultures, but this view that an individual's own culture is superior to all the rest is
ordinal variables
allow us to rank order attributes with regard to which has less versus which has more of a characteristic - Both qualitative and quantities - Qualitative bc people are placed into categories used to represent characteristics - These categories have real numerical associated so also quantitative - Logical order, important to note that the research not able to make comparison regarding the size of difference between categories - 3 primary properties for categorizing: o Mutually exclusive o Participant can only "fit" into one of categories o Researcher must follow a logical ordering of categories (ordinal variable of military rank) o Each category must be balanced to represent amount of a characteristic possessed by participant - To quantify ordinal level data for statistical analysis, numerical values are assigned to represent each category - The numbers sued to represent each quality of variable must increase such that higher numbers used to represent high values - Interval between each category don't represent equal distances - Point zero does not exist
data
any record or observation; examples of data include the number of years employed, grade point average, annual income, gender, number of hours spent doing hw
communication apprehension
defined as the fear or anxiety associated with either real or anticipated communication with another person or persons - (paper on first day) - 1st CA is about fear and anxiety; people exhibit fear and anxiety differently; some people are more fearful and anxious about almost everything, whereas others are more focused in their fear and anxiety - People who are said to be high CA will experience fear and anxiety in every context when having to communication; do not just stop having anxiety just bc they don't have to communicate at that moment - happen even when they think about it; experience severe anxiety while practicing a public speech alone in their house, or thinking about interacting with someone the next day - Low CA - higher talker and actively seek opportunities to interact with other; few situations cause you to be fearful or anxious
std deviation
directly related to the variance of a sample - Stems from sum of squares - On average tells us how far each score differs from the average score
subscale
hypothetical subdivision of mental measure
what are the 4 units of analysis?
individuals, dyads, groups, organizations
parameter
is the real average of communication apprehension in a population, and a statistic is the average of communication apprehension within a sample
descriptive stats
is used to organize and summarize information or data o Allows a researcher to provide a description of what actually "exists" in the data o Most commonly used percentages o Could use to summarize students in a particular class o Summarize data, or describes data
variance
measures how wide or spread out a distribution is, or the average distance of the score for an interval or ratio scale from the mean in squared units - Variance dependent on the sum of squares, we know that as our scores become more different, the variance will increase (and the opposite is true, as the scores become more similar the variance will decrease)
extraneous variables
ny variables that you are not intentionally studying in your experiment or test. When you run an experiment, you're looking to see if one variable (the independent variable) has an effect on another variable (the dependent variable). ... These undesirable variables are called extraneous variables
- How to determine the independent variable and dependent variable?
o Consider the following hypothesis from Web (2003) study on the effect of a student interest in a topic on recall of information: H: Lectures utilizing interest-based examples should result in participants with higher scores on subsequent tests of cognitive recall o Independent variable = lecture examples o Lectures were manipulated by showing students 2 diff versions of video tapes: (1) lectures that incorporated current examples and (2) lectures using standard examples o Cognitive recall of information = dependent variable o Researchers are trying to example the difference in information recall as a result of the types of examples used in lectures - Some studies examine variables that the researcher does not manipulate; rather the difference in a variable may occur naturally - Consider the following research question from a 1995 study examining communication satisfaction of children from divorced families: o H: communication satisfaction with parents will be lower amount child from divorced families than among children from intact families Family status (divorced or intact) = dependent variable Goal is to determine whether the differences in family status affect the dependent variable or communication satisfaction - RQ: what is the relationship between parents who engage in verbally aggressive behavior at sporting events and a child's verbal aggressiveness? o We simply want to ascertain a possible relationship between 2 variable o In other words, there is not a dependent or independent variable o If we said that a parent's level of verbally aggressive behavior at sporting events caused his or her child level of verbal aggression, we could say the parents level of verbal aggression is the independent variable and use that in attempt to predict child's level of verbal aggression
generalized belief scale
o Created by McCrisken (1966) as a tool for examining the strength of an individual's belief on a given topic o A belief - concerns our perception of reality about whether something is true or false
dynamic/consistent
o Dynamic aspects - the degree to which a personality trait can change over time or during a certain contexts or situations o Consistent aspects - the degree to which a personality trait doesn't change overtime or during certain contexts or situation o Continuum of personality traits ranging from consistent to dynamic o Many external and internal factors can cause personality traits to change o Ex: moving to another country, experience a traumatic event o Many aspects of our individual personalities do not alter over time
personality and communication
o Personality: the total psychological makeup of an individual, which is a reflection of her experiences, motivations, attitudes, beliefs, values, and behaviors derived from the interaction of these elements with the environment external to the individual o Personality can be defined as the sum of an individual's characteristics that make them unique o Personality has 2 basic dimensions that are important to communication researchers: external/ internal and dynamic/consistent
external/internal
o The 1st dimension of personality distinguishes the external level - the observable characteristics of an individual from the internal level - the individual's attitudes, values, beliefs, interest, aspirations, and motivation o Often use external personality to determine internal personality during interpersonal reactions o Expect people to behave according to the way they think or feel but not always consistent
differences in kind
occur when 2 or more groups do different things associated with their groups - Example: football players play football and cheerleader lead the crowd in cheers - These 2 groups are fundamentally different and have different behaviors - Don't want to see a 90lb cheerleader get tackled by 250 lb football player
infernteial stats
once data is described in terms of percentages or averages, the data can be further analyzed to make inferences from the smaller group of data (known as the sample) to a larger group of data (referred to as the population) o Used when a researcher wants to make predictions o Sample: subset of larger population, or all the possible persons who could fall into a particular category
what are the common interval variables
personality and communication, interal and external dynamic and consistent
germinal idea
refers to the spark that causes individual to realize that something new can be measured
conceptualization
the development and clarification of concepts of your germinal idea
perceived knowledge
understanding what a person knows is important, it is also important to be able to measure what a person thinks they know
Statistic:
when we calculate information from a sample of data
