Types of Statistical Analyses Matrix
Carrington Sherman, Donna Crawford, Henry Izeke, Stella Crozier, Trish Gordon
HCS 542
April 2, 2018
Lane Baggett
Purpose
Example of when it would be used
provides information about significance of differences between groups
provides information about significance of relationships between variables
provides information about a single sample vs. two or more samples
is parametric or nonparametric
Descriptive Statistics: Mean
The average
Average grades in the class
Mean is different from the median because it is the sum of the data set then divided by the total of the number of the data set. So, the average.
Mean and median are both a type of average
A single sample is getting average of the data set
Two or more samples you are getting mean of all means or the average of all the samples
Parametric
Descriptive Statistics: Median
Midpoint in a data set
Median salary; Middle of a groups salary which different from the mean because low and high salaries could way the average.
The median is different from the mean because it is the midpoint of the data set or sample being used. The mean and median can be the same but rarely
Median and mean are both a type of average
With one set or sample it is simply the midpoint such as 3, 5, 12; 5 is your midpoint.
If there is two sets or samples put all the numbers together and then find the midpoint.
Parametric
Descriptive Statistics: Mode
Value that appears most often
Looking at bar chart of incomes that is mostly around minimum wage then you have a couple that make a million dollars.
Mode shows the top number or numbers that are most used but it can affect both the mean and median.
Mode can affect both the median and the mean
There can be no mode
One mode or numbers: unimodal
Two modes or numbers: bimodal
Three modes or numbers: trim (Smith, 2018)odal
Parametric
Descriptive Statistics: Ratio Variables
Ratio variables are used to show comparisons between two or more samples. They are characterized by having answers that are numbers on a scale; the difference between two samples is expressed as having a numeric value which is significant and where a zero response indicates that there is none of that variable, also known as having a true zero.
Used for measurements such as height, weight and BP.
The significance of differences between groups is expressed as a ratio of two measurements that are being compared. For example, a weight of 4 grams is twice the weight of 2 grams, or a 2 to 1 ratio.
Used to show comparison between two or more variables. Can be used to compute the following: how frequent one variable occurs (counts) compared to another, mean, median, mode, percentiles of variables, add or subtract, standard deviation, and ratio. The difference between variables can be quantified. There is an order to the values on the scale.
One sample demonstrates what value is present. Two or more samples demonstrate comparison.
Parametric
Descriptive Stati ...
Types of Statistical Analyses MatrixCarrington Sherman, Donn.docx
1. Types of Statistical Analyses Matrix
Carrington Sherman, Donna Crawford, Henry Izeke, Stella
Crozier, Trish Gordon
HCS 542
April 2, 2018
Lane Baggett
Purpose
Example of when it would be used
provides information about significance of differences between
groups
provides information about significance of relationships
between variables
provides information about a single sample vs. two or more
samples
is parametric or nonparametric
Descriptive Statistics: Mean
The average
Average grades in the class
Mean is different from the median because it is the sum of the
data set then divided by the total of the number of the data set.
So, the average.
Mean and median are both a type of average
A single sample is getting average of the data set
Two or more samples you are getting mean of all means or the
average of all the samples
Parametric
Descriptive Statistics: Median
Midpoint in a data set
2. Median salary; Middle of a groups salary which different from
the mean because low and high salaries could way the average.
The median is different from the mean because it is the
midpoint of the data set or sample being used. The mean and
median can be the same but rarely
Median and mean are both a type of average
With one set or sample it is simply the midpoint such as 3, 5,
12; 5 is your midpoint.
If there is two sets or samples put all the numbers together and
then find the midpoint.
Parametric
Descriptive Statistics: Mode
Value that appears most often
Looking at bar chart of incomes that is mostly around minimum
wage then you have a couple that make a million dollars.
Mode shows the top number or numbers that are most used but
it can affect both the mean and median.
Mode can affect both the median and the mean
There can be no mode
One mode or numbers: unimodal
Two modes or numbers: bimodal
Three modes or numbers: trim (Smith, 2018)odal
Parametric
Descriptive Statistics: Ratio Variables
Ratio variables are used to show comparisons between two or
more samples. They are characterized by having answers that
are numbers on a scale; the difference between two samples is
expressed as having a numeric value which is significant and
where a zero response indicates that there is none of that
variable, also known as having a true zero.
Used for measurements such as height, weight and BP.
The significance of differences between groups is expressed as
a ratio of two measurements that are being compared. For
example, a weight of 4 grams is twice the weight of 2 grams, or
3. a 2 to 1 ratio.
Used to show comparison between two or more variables. Can
be used to compute the following: how frequent one variable
occurs (counts) compared to another, mean, median, mode,
percentiles of variables, add or subtract, standard deviation, and
ratio. The difference between variables can be quantified. There
is an order to the values on the scale.
One sample demonstrates what value is present. Two or more
samples demonstrate comparison.
Parametric
Descriptive Statistics:
Interval variables
Used for showing comparisons between samples. Characterized
by having answers that are numbers on a scale where the
difference between two values is significant but where zero
does not stand for having no values or data for that category of
data.
Temperature measurements in Celsius or Fahrenheit. Zero
temperature does not indicate absence of heat.
The significance of differences between groups is expressed
by comparing the frequency a value occurs in one group
compared to the frequency that same value occurs in another
group, or the mean median or mode for each group can be
compared.
Can be used to compute the following: how frequent one value
occurs (counts) compared to another, mean, median, mode,
percentiles, add or subtract, standard deviation. The difference
between variables can be quantified. There is an order to the
values on the scale. The difference between two or more
variables can be computed, such as 40 degrees Fahrenheit is 30
degrees higher than 10 degrees Fahrenheit; but 40 degrees
Fahrenheit is not twice as hot as 20 degrees Fahrenheit.
One sample demonstrates the value that is present. Two or
more samples demonstrate comparison.
Parametric
Descriptive Statistics:
4. Ordinal variables
Also called ranked variables, is an ordered series of responses
where the order matters but not the difference between the
values; such as from smallest to largest, best to worst, or 1 to
10 scale.
Used for assessing patient’s pain level using a scale from 1 to
10, where 1 is associated with no pain and 10 is the worst pain
the patient has ever experienced. Pain level of 7 is more painful
than a pain level of 4.
Severity of illness is another example.
Can be used to compare responses between groups through
comparing the frequency a value occurs in one group versus
another or by computing the median or mode of one group and
comparing to another group.
Can be used to show comparison between variables to compute
how frequent a value occurs, median, mode and percentiles.
There is an order to the values on the scale.
A single sample would demonstrate which value on the scale is
present. Two or more samples demonstrate comparison.
Non-parametric
Descriptive Statistics:
Nominal variables
Represent groups with no obvious rank or order. Often used to
label variables that do not have quantitative value. Also can be
considered a label.
Give a value to a description in a survey, such as; what is your
hair color
1-Brown
2-black
3-blonde
4-red
So that it is easier to analyze the data.
Nominal variables are different because it is used as a label not
necessarily the actual response or data itself.
Nominal is a type of variable that can be used with a
quantitative data set, similar to interval, ordinal and continuous
5. variables
This couldn’t be a single sample, it would need to be two or
more, as it is for groups with no rank or order
Parametric
Descriptive Statistics:
Binomial variables
A sub category of categorical variables, when only two choices
are possible, either yes or no.
A coin toss, with only two possible outcomes heads or tails.
The statistic would show, the number of times the coin was
tossed, and the number of times it lands on heads, and the
number of times it lands on tails.
Binomial is different because there are only two possible
options, not infinite such as with continuous
Binominal is similar because it is used to help take qualitative
data and apply it to quantitative capable analysis. Making it
similar to the other variables.
This could be either single or two or more because it represents
variables that only have two possible outcomes, so if the
outcome is the same every time it could be a single sample.
Parametric
Descriptive Statistics:
Continuous variables
Any value within a range or numbers, meaning the value is not
rounded to the nearest value, it can be fraction or long decimal.
Recording a person’s weight, it could be 180.00 or it could be
180.0010
Another example is age, the number could be 33years or it could
be 33 years 20 days and 3 hours
Continuous is different because there are never ending
possibilities for its value, unlike the other variables that have a
defined exact value.
Continuous is similar because it is a variable used in
quantitative data analysis like the other variables.
This could be a single sample, or two or more because it
represents a value with never ending value possibilities, but it
6. could be one value or a group of values.
Nonparametric
Descriptive Statistics:
Discrete variables
Henry
Inferential statistics: t-tests
Henry
Inferential statistics: ANOVA
Henry
Inferential statistics: regression analyses
Carrington
Inferential statistics: various correlational analyses
Carrington
7. Inferential statistics: chi-square
Carrington
References
Jacobsen, K. H. (2017). Introduction to health research
methods. A practical guide (2nd ed.). Sudbury, MA: Jones &
Bartlett.
Neutens, J., & Rubison, L. (2014). Research techniques for the
health sciences (5th ed.). San Francisco, CA: Pearson Education
Smith, J. (2018). How Do People Use Mode, Mean & Average
Everyday? Sciencing.