92 terms

variable

the changing characteristic being measured

probability (p)

how likely it is for an outcome to occur

statistic

an estimate derived from a sample

ex: mean age of students in a class

ex: mean age of students in a class

parameter

an estimate for the whole group (population)

ex: mean age of students in the US

ex: mean age of students in the US

population (N)

group of objects or people that are alike

(determined by research)

(determined by research)

sample (n)

a group of cases selected from the population for research

(population -- parameter)

(sample --statistic)

(population -- parameter)

(sample --statistic)

discrete variable

finite number of values; categories or counting units of whole numbers

ex: number of children in a family, types of insulin used

ex: number of children in a family, types of insulin used

dichotomous variable

type of discrete variable with only two categories

ex: yes/no; male/female

ex: yes/no; male/female

continuous variable

infinite number of values between any two values, equal intervals, or any level of data capable of having a decimal

ex: height, weight, time in minutes, temperature

ex: height, weight, time in minutes, temperature

extraneous variable

variables that we may/may not understand; can influence findings

demographic

an example of an extraneous variable that is intrinsic to subjects or that cannot be changed due to research situation

independent variable (x)

variable measured or controlled by researcher; what you think is affecting the outcome

dependent variable (y)

outcome/final result; variable you wish to change due to experimental treatment

nominal data

lowest level of measurement; the numbers are just used as names showing sameness or differentness of a particular quality; discrete

ordinal data

we can rank things; discrete

ex: mild, moderate, or severe

ex: mild, moderate, or severe

interval/ratio data

highest level of measurement; exhaustive, exclusive, and ordered with numerically equal intervals

ex: temperature, scores on a test

ex: temperature, scores on a test

visual analog scale

usually reported in categories or can be averaged

ex: rate your pain on a scale from 0-10

ex: rate your pain on a scale from 0-10

Likert-type scale

usually total and provide an average

ex: SA, A, U, D, SD

ex: SA, A, U, D, SD

semantic differential scale

participants indicate their feelings about statistics in reference to the two opposing adjectives

ex: good/bad, smart/impaired, strong/weak

ex: good/bad, smart/impaired, strong/weak

frequency distribution

frequency of each measure of a variable; summary of the numerical counts of the values or categories of a measurement

cumulative frequency

cumulative measure of frequency of a variable; number of observations with a value less than the maximum value of the variable interval

grouped frequency

a frequency distribution with the distinct intervals or groups to simplify the information; used when the entire frequency distribution would be too large to be meaningful; drawback is that some data may be lost

frequency table

presents a big-picture of your data; presents the values of the dependent variable, from lowest to highest, with a count of the frequency

(ie: how often each value occurred)

(ie: how often each value occurred)

cumulative percentage distribution

summing of percentages from the first category of the table, ending with a cumulative percentage of 100%; same idea as cumulative frequency, but expressed as a percentage

quantile

diving data into different portions or bins

percentile

diving data into 100 equal parts

25% and 75% are quantiles

50% of the subjects will fall between the 25% and 75% because everything is equally divided

25% and 75% are quantiles

50% of the subjects will fall between the 25% and 75% because everything is equally divided

interquartile range

50% of the subjects will fall between the 25% and 75% because everything is equally divided

ordinal

interval ratio

interval ratio

percentiles can be used with what types of data?

quartile

diving the data into four equal parts

nominal

ordinal

ordinal

what type of data can be used in bar charts?

bar charts

lines do not touch; each answer is distinct and in no particular order

x-axis = nominal variables

y-axis= frequencies or percentages

highest variable = mode

x-axis = nominal variables

y-axis= frequencies or percentages

highest variable = mode

ordinal

continuous

continuous

what type of data can be used in histograms?

histograms

shows the flow of data in ranked order; bars do touch

x-axis = ordinal or continuous data

y-axis = frequencies or percentages

highest variable = mode

x-axis = ordinal or continuous data

y-axis = frequencies or percentages

highest variable = mode

continuous data

what type of data can be used in line graphs?

line graph

a figure that is developed by joining a series of points with a line to show how a continuous variable changes over time

scatter plot

each dot is subject and is placed where the score for variable x and the score for variable y are located

positive = upward slope

negative = downward slope

weak = scattered

moderate to strong = closely spaced dots creating a line

number of dots = sample size

positive = upward slope

negative = downward slope

weak = scattered

moderate to strong = closely spaced dots creating a line

number of dots = sample size

outliers

can be seen on a scatter plot; can be an interest especially if most of your data has a strong correlation and the outliers are extreme

descriptive statistics

enable the researcher to describe the data; the level of measurement determines what analyses are possible; analysis often begins with central tendency

central tendency

mean, median, and mode; the best one to use depends on the level of measurement

mode

can be used for nominal, ordinal, or interval/ratio data; it is the only option for nominal

modality

number of peaks in a curve

unimodal

one peak

bimodal

two peaks

trimodal

three peaks

median

can be used with ordinal or interval/ratio data

mean

can be used with interval/ratio data

standard deviation

the average distance the values in your distribution fall from the mean

best measure of central tendency in interval/ratio

best measure of central tendency in interval/ratio

one standard deviation

68%

two standard deviation

95%

three standard deviations

99%

more variability from a large range of scores

large standard deviation indicates:

less variability from a small range of scores

small standard deviation indicates:

population mean

mu (lowercase)

population standard deviation

sigma (lowercase)

range

simplest measure of deviation; SD is larger when the range is larger

ordinal

interval/ratio

interval/ratio

what type of data can be used in ranges?

z scores

measure the distance between the mean and an observation

right

increasing the mean will shift the curve to the....

left

decreasing the mean will shift the curve to the...

flattent the curve

increasing variance (SD) will....

heighten the curve

decreasing the variance (SD) will.....

inferential stats

the ability to make inferences about populations based on sample measurements; involves associating probability with each variable

probability distribution

may be estimated by the normal distribution; uniform and never change

population

a group of objects or people that is alike on one or more dimensions as defined by the researcher

target population

the entire population in which the researcher is interested and would like to generalize the results of a study

accessible population

the population of subjects available for a particular study; subset of the target population

sample

a group of cases selected from a population for the purpose of conducting your research

representative sample

contains all the attributes of the population in the same proportion that they occur in the population

inclusion criteria

required characteristics to be included in the sample based on characteristics of the population

exclusion criteria

factors which eliminate a subject from the sample

generalizability

based on findings from the sample, we make "inferences" or statements of what we think is true for the population

enhances the generalizability of findings

supports inferences for the population

supports inferences for the population

having a representative sample:

sampling method

process researches use to select subjects from the population being studied

random sampling

every member of the population has the same chance of being selected and the probability of being selected is known

simple random sampling

not feasible with large populations because the researcher has to have access to every member of the population; would work to draw a sample of students or RNs at a hospital

systemic random sampling

selecting subjects based on a standardized rule

ex: numbering the population, determining a starting point, and selecting every 4th person

not feasible with large populations

ex: numbering the population, determining a starting point, and selecting every 4th person

not feasible with large populations

stratified random sampling

the population is divided into subpopulations (strata) based on characteristics of interest. the sample is then randomly selected from the subpopulations

cluster (random) sampling

uses a group, unit, or cluster rather than an individual; used when it is difficult to locate a list of the entire population

nonprobability sampling

subjects do not have the same chance of being selected for participation; it is not randomized

convenience sampling

most popular form of non probability sampling in healthcare research; subjects are in the right place at the right time; quantitative

quota sampling

select proportions of the sample for different subgroups; quantitative

network sampling

utilizes social networks to gather information; frequently used in groups that hesitate to participate in research; qualitative

purpose sampling

subjects selected because they have particularly strong bases of information

sampling error

differences between the sample and the population that occur due to chance

sampling bias

systemic error made in the sample selection that results in a nonrandom sample and therefor the findings are not representative of the population of interest

sampling distribution

all the possible values of a statistic from all the possible samples in a given population

central limit theorem

essentially with more and more samples. the resulting distribution of the averages tend to look more and more bell-shaped and normally distributed

null hypothesis (H0)

there is no relationship, no association, no effect, and no difference; we must presume the null is true until we can disprove it

alternative hypothesis (H1)

what the researcher actually believes to be true or present; there is a significant difference or a signifiant relationship

directional hypothesis

tells the specific direction of a relationship or in what way the populations will differ

non-directional hypothesis

simply tells whether there is a relationship or a difference

...

...