Home > Preview
The flashcards below were created by user
lboomer
on FreezingBlue Flashcards.

Descriptive Statistics
Descriptive statistics tell what is. There are 9 in class. Theyare used to organize and summarize information or data. They allow a researcher to provide a description of what actually exists in the data. Researchers use descriptive statistics to help us reduce large amounts of data to a more manageable size. Perhaps one of the most commonly used descriptive statistic is percentages.

Inferential statistics
Are often the next stage is data analysis. Are used when a researcher wants to make predictions. If I look at exam scores of 3 students, can I predict the total classes scores. More unknowns= more errors.They use a smaller group, or a sample to compare data to larger groups.

Error
 Type 1: Alpha rejecting the null hypothesis when the null is true. "there is some relationship" but there isnt.
 Type 2: Beta You keep the null hypothesis but it is false. "Theres no relationship" but there is.

Sample:
Drawn from populations. Smaller than populations. Take 10% of the 23 comm 300 students.

Population
an entire set of objects, observations,or scores that have some characteristic in common.

Statistics
The average of communication apprehension within a sample. They allow us to summarize or describe data. They enable scientists to make predictions and make sense of the world.

Measures of central tendency:
Mean, Median, Mode



Mode:The value that occurs most in a set of numbers

Measures of variability:
range, sum of squares, variance, standard deviation.

Range
Highest number subtracted by lowest number

^{Sum} of squares
SS= ex2 (ex)2/N


Histogram
Useful for spotting trends. XAxis= horizontal<> YAxis= Vertical ^

Standard Deviation:
SD= SS/N1(square rooted)

zscore
x minus the mean divided by standard deviation
xmean/SD

Frequency distribution
how scores scatter on a plot

Skewness
If something is off center, it is skewed. Two types: Positive, Negative.

Positive Skew:
a curve in which the tail of the curve is longer on the right side of the distribution. The mean will be greater than the mode.

Negative Skew
A curve in which the tail is longer on the left side. The mean will be less than the median or mode.

Kurtosis
Deals with the shape of the curve. Tall and narrow, or short and squatty.

Platykurtic kurtosis
Short and squatty.... looks like a tea pot... wide range of variability in x axis, low range in y axis

Leptokurtic kurtosis
Looks like a tall wine bottle.... low x axis, wide range of y axis.... tall and narrow.

Mesokurtic kurtosis
in the middle... normal... roughly equal variances in both y and x axis.

Measurement levels
Nominal, Ordinal, interval, ratio

Nominal
Nominal measurement is a simple classification. Numbers are selected and meaningless. Lowest level of measurement.

Ordinal
Rankorder measurement. Assigns numbers to things in such a way as to reflect relationships among things. Allows us to see hierarchal levels between groups.

Interval
Interval measurement identifies the distance between any two things that are measured. We assign numbers to things in such ways that the distance between any two things is measured. They are probably the most common measures used by scientists.

Ratio
A ratio scale has all of the characteristics of an interval scale, but in addition it has a true zero point as its origin. In ratio measurements, zero has an absolute value of its own.

germinl idea
a spark, a lightbulb of an idea for research... makes researcher start wondering if it can be researched.

Conceptualization and Operationalization
 Conceptualization Clearly defining your variables.
 Operationalization: Process of measuring what you conceptualize. A good operationalization is like a recipe card. It tells you a very detailed procedure and what to do and when to do it.

parametric statistical test vs nonparametric statistical test
 parametric researchers can calculate measures. Parameter = value of population.
 nonparametric make no assumptions about parameters of the popultion from which the research sample can be drawn.

Measurement:
the process of systematic observation and assignment of numbers to phenomena according to rules.

Isomorphism
means identity or similarity of form.

Likert Scale:
Disagree, Strongly disagree, agree, etc etc 5 terms.... ordinal

Hypothetical Variable
A variable that a researcher cannot directly observe, but is inferred from other variables that are observable and measured directly.

Reliability:
Stability and consistency over time. The accuracy that a measure has in producing stable, consistent measurement.

Scalar Reliability:
The reliability of individual research scales...... Good 1,2,3,4,5,6,7 Bad(circle whihc one fits best to individual personality)

TestRetest Reliability
Testing the same thing on more than one occasion to make sure you get the same data.

Alternate forms reliability
Same thing as test retest, but you use each test as its own separate form of reliability.

Split half reliability
Involves computing two scores for each participant on the basis of one administration of the test. One of the scores comes from one half of the test and the other score comes from the other half.

Hoyt Analysis of variance reliability
internal reliability estimates. Another method of split halves.

Cronbachs alpha reliability
1,2,3,4,5 are flipped on the results... 5= good, 1=bad

validity
the degree to which the instrument measures what it is intended to measure.

approaches to validity:
face or content , criterion, predictive, concurrent, retrospective, construct or factual

threats to validity
inadequate preoperational explication of concepts, monooperation bias, interaction of different treatments, interaction of testing and timing, restricted generalizability accross constructs.

improving reliability:
item construction, length of the instrument, administration of the test.

Survey
A social scientific method for gathering quantifiable information about a specific group of people by asking the group members questions about their individual attitudes, values, beliefs, behaviors, knowledge, and perceptions.

Questionaire
A form containing a series of questions and mental measures that is given to a group of people in an attempt to gain statistical information about the group as part of a survey

Descriptive survey
is designed to find out how common a phenomenon is within a group of people. Ex: a researcher may be interested in communication apprehension accross the United States.... US is the group of people...

Analytical Survey
the purpose of this survey is to explain why people think or act as they do by identifying likely causal influences on their attitudes or behavior.

interview schedule
 the list of survey questions an interviewer reads an interviewee when conducting an oral survey...
 Paper and pencil methods= questionairre
 oral interviewing= interview schedule

5 steps to conducting a survey
 1) pick your questions
 2) create clear instructions
 3) study design
 4)Data processing and analysis
 5) Pilot testing

Question formats
Nominal, ordinal, interval, ratio, open ended)

Disseminating(distributing) surveys
 2 ways: interviewing and selfadministration
 Examples: face to face interviews, telephone interviews, mass administration, malied administration, internet administration.

Problems with survey research
Response rate, unit nonresponse(failure to obtain any survey measurements on a sample unit), item nonresponse(occurs when an individual participant fails to answer individual or groups of questions on a survey).... ALL CAN BE IMPROVED WITH CLEAR AND CONCISE DIRECTIONS,MAKING THE SURVEY EASY TO FILL OUT, KEEPING IT SHORT, A GOOD COVER LETTER...

When to use a survey: 4 questions
 1) Do you know what you want to ask?
 2) Do you need to collect data?
 3) Does your audience know any information youre looking for?
 4) Is your goal generalizability? (is it driven towards a population or a sample?)

Pilot test
a small scale test run for a planned piece of empirical research... They are used to make sure instructions are easy to follow and survey questions make sense to participants... When pilot testing: Use actual survey population members, anticipate survey context, test parts of the survey, determin a pilot sample size, and ask questions after someone completes the survey...

Cross sectional survey,
longitudinal survey design:1)trend design, 2)panel design,
accelerated longitudinal design
 Cross sectional survey is used when a researcher wants to get information from a group of participants at a given point in time.
 longitudinal survey design it allows you to make statements about variable order. 1st type: trend design used to examine different samples of people at different points in time. 2nd type: Panel Design When a researcher recruits a series of participants who agree to be surveyed periodically over a given amount of time.
 Accelerated longitudinal design when a researcher wants to see how things change over a long period of time during a short period of time.

Threats to validity
 1) history events in the participants lives that occur during the experiment... ex: 9/11
 2) maturation any psychological or physiological changes taking place when the participants that occur with the passing of time regardless of the experimental manipulation
 3)testing effects changes in what is being measured brought about by reactions to the process of measurement.
 4)instrumentationunwanted changes in characteristics of the measuring instrument or in the measurement procedure.
 5) attritionlosing participants in an experiment.

PRE EXPERIMENTAL DESIGNS
one shot case study, one group pretest postest design, static group comparison,

True Experimental Designs
the pretest postest control group design, postest only control group design, solomon four group design, factorial designs, quasiexperimental designs, nonequivalent control group designs.

random assignment
selecting a random group of people throughout a table of numbers

independant variable
what the experimentor changes in the the experiment

dependant variable
what remains unchanged by the experimentor

control
making sure the experiment is conducted as is... meaning there are no external factors messing with the results.... everything is controlled.

