-
research
systemmatic study of one or more problems usually posed as research questions germane to a specific dicipline.
-
population
larger group researcher wants to draw conclusions about.
-
parameter
- characteristic of a population.
- usually unknown but are estimated with statistics.
-
sample
group of the population that is actually studied.
-
statistic
characteristic of a sample.
-
statistics
a branch of applied mathematics that deals with collecting, organizing and interpreting data using well defined procedures.
-
purpose of statistics (3 parts)
- describe and summarize info, reducing it to smaller, more meaningful data sets.
- make predictions or generalize about occurnaces based on observations.
- identify associations, relationships or differences in observations.
-
types of statistics (2)
descriptive and inferential.
-
descriptive statistics
characterizes data by summarizing it into more understandable terms without losing or distorting much information.
-
inferential statistics
provides predicitions about a population's characteristics based on information from a sample of that population.
-
data
raw materials of research gathered from a sample that has been selected from a population.
-
variable
- characteristic being measured that varies among persons, events, or objects being studied.
- a concept that a method of measurement has been determined for.
-
measurement
assignment of numerals to objects or events according to a set of rules.
-
types of measurement scales (4)
nominal, ordinal, interval & ratio.
-
nominal scale
- lowest form of data.
- organizes data into discrete untis.
- allows researcher to assign numbers that classify characteristics of people, objects or events into categories.
- assignment of numerals is arbitrary.
-
qualitative nominal variable
-
categorical nominal variable
-
types of nominal variables
categorical and qualitative
-
ordinal scale
- places characteristics into categories and categories are ordered in some meaningful way.
- distance between categories is unknown.
-
interval scale
- distances between category values are equal due to some accepted physical unit of measurement.
- i.e. F - temperature
-
types of interval variables
continuous and discrete
-
continuous interval variable
may take on any numerical value within a variable's range.
-
discrete interval variable
takes on only a finite number of value between two points.
-
ratio scale
- most precise level of measurement.
- meaningfully ordered characteristics with equal intervals between them and the presence of a zero point that is determined by nature.
- i.e. pulse, bp, weight
-
meaningfullness
clincial or substantive meaning of the results of statistical analysis.
-
davidson's principles
- principles of statistical data handling to fill the gap between getting data into the computer and running statistical tests.
- summarize the key dilemas that researchers face when entering data into the computer.
- he wrote a book in 1996 about them. our book only covers 18.
-
dp - appropriate data principle
- you cannot analyze what you do not measure.
- must anticipate variables needed to expalin results.
-
dp - social consequences principle
- data about people are about people.
- can have social consquences.
- i.e. drug proven not to work better, unethical to advise people to take it?
-
dp - data control principle
- take control of structure and flow of data.
- monitor procedure for layout of data record.
-
dp - data efficiency prinicple
be efficient in getting your data into a computer, but not at the cost of losing cucial information.
-
dp - change awareness principle
data entry is an interactive process. try to use the computer to do as much computing and debugging as possible.
-
dp - data manipulation prinicple
- let the computer to do as much work as possible.
- let it manipulate the data for you by instructing it to do so.
-
dp - original data principle
always save a computer file of the original, unaltered data.
-
dp - default prinicple
- know your software's defualt settings and whether they meet your needs.
- (especially concerning missing values)
-
dp - complex data structure principle
if your software can accommodate complex data structures, then you might benefit from using that software feature.
-
dp - software's data relations principle
know if your software can perform the following four relations and if so, what commands are necessary for it to do so: subsetting, catenation, merging and relational database construction.
-
dp - software's sorting principle
know how to perform a sort in your software and whether your software requires a sort before a by group analysis or before merging.
-
dp - impossibility/ implausibility principle
use the computer to check for impossible and implausible data.
-
dp - burnstein's data sensibility principle
run your data all the way through to the final computer analysis and ask yourself whether the results make sense.
-
dp - extant error principle
data bugs exist even if you've corrected mistakes it's possible you've missed something.
-
dp - manual check principle
- nothing can replace another pair of eyes to check over a data set.
- check it yourself or get someone else to do it.
-
dp - error typology principle
- debugging includes detection and correction of errors.
- try to classify each error as you uncover it.
-
dp - kludge principle
- sometimes the way to manipulate data is not elegant and seems to waste computer resources.
- patching together cpomputer demands awkwardly to make data do what you want.
-
dp - atomicity principle
- you cannot measure below the data level that you observe.
- i.e. age 21-25 nominal (lowest)
- age 26-29
- vs. age? ___ more precise
-
quantitative research
using specific methods to advance the science base of the discipline by studying phenomena relevant to the goals of that discipline.
-
individuals
objects being described by a set of data.
-
quantitative research methods
experiments, surveys, correlational studies, meta-analysis, and psychometric evaluations.
-
bar chart
- simplest foorm of a chart for nominal or ordinal data.
- category labels horizontally in a systematic order with vertical bars with spaces between.
-
histogram
- appropriate for interval, ratio, and sometimes ordinal variables.
- similar to bar chart except bars are placed side by side.
-
polygon
- a chart for interval or ratio variables.
- it is equivalent to a histogram but appears smoother made by connecting midpoints of the top of each bar.
-
pie chart
a circle that has been partitioned into percentage distribution of quantitative variable total area 100% = 360 degrees.
-
statistical table
data is organized into values or categories and then described with titles and captions.
-
working table
- a frequency distribution for interval or ratio variables.
- an ordered array of values.
-
mean and formula
- best known and widely used average.
- the center of a frequency distribution.
- x bar = M = sum of x/n
-
measures of central tendency
- mean
- median
- mode
- - center of trend or average
-
median
- middle value of a set of ordered numbers.
- point where 50% of distribution falls below and above.
- not affected by outliers.
- - place #'s in order and the middle number is median if n is odd
- - if n is even average the middle two #'s
-
mode
- most frequent value or category in a distribution.
- not calculated, just observed.
- if all scores are different then there is no mode.
- -use when dealing with frequency distribution for nominal data.
-
homogeneous
- having low variability.
- numbers clustered.
-
heterogeneous
- having high variability.
- numbers are spread out.
-
variability
- measure of spread or dispersion.
- measure of degree to which scores in a distribution are spread out or clustered together.
-
types of variability
- standard deviation (SD)
- range
- interquartile range
-
interquartile range
- range of values extending from the 25th percentile to the 75th percentile.
- - divide by 2 for the semi interquartile range
-
range
- max-min... highest #- lowest #.
- simplest measure of variability.
- sensitive to extreme #
- unstable since it's only based on two numbers.
-
standard deviation (SD)
- measure of dispersion of scores around the mean.
- most widely reported, indicates spread.
- low SD means close together and high means spread out.
-
interpercentile measures
- interquartile range (IQR)
- -range of values from P25 to P75
- not sensitive to outliers.
- used on growth charts.
- good with skewed data.
- use with median.
-
skewness
- non symmetrical distribution.
- measure of symmetry.
-
kurtosis
measure of flatness.
-
measures of variability
- standard deviation
- range
- interpecentile measures
-
types of data transformation
- square root transformation
- log transformation
- inverse transformation
-
-
-
square root transformation
-
-
-
-
-
pearson's skewness coefficient
-
fisher's measure of skewness
-
-
-
-
-
-
-
-
-
|
|