Descriptive Study Designs
Case Report |
Case Series |
Cross sectional Surveys |
Ecologic study |
A report of one individual which descriptive research is written. |
A small group of people who have similar diagnosis |
A survey that is done in a short period of time and its focus is an individual |
Comparing variables when the unit of analysis is aggregated data |
This qualitiative information is in chronological order |
Descriptive information about research on the groups |
There is control with population and measurements |
Can help evaluate policies, rules, or programs |
There is information on only one person |
The information is only about the small group |
Rare conditions are difficult to survey and could be response bias |
Possibility of confounding factors |
|
|
Types of data
Nominal data |
Ordinal data |
Continoues data |
Discrete data |
The order is not intrinsic and the difference between level is meaningless |
There is an order made among categories |
In a range there can be any value |
the values are integers with fixed amounts |
gender, race, ethnicity |
ranges, stage of cancer |
age, weight, temperature |
number of meals eaten in three days |
Tables and Charts
The most simple table is frequency distribution is a summary of frequencies. Relative frequency is dividing the number of people in each group by total number of people. |
Bar charts |
spot map |
box plot |
histogram |
two way scatterplot |
line graph |
steam and leaf plot |
area map |
|
|
Ratios and Rates
A ratio is two values that are compared and it is calculated by dividing the numerator and denominator then multiplying 10^z (0,1,2,3,4,5 which equals 1,10,100,1,000,10,000,100,000). Rate is a measure of frequency in which a health related outcome occurs in a short period of time. Incidence rate is the number of new cases occurring in a given time. Prevalence rate is the frequency of existing cases at a given period of time. Point-prevalence is the proportion of a health related outcome at a point in time. An attack rate is when new cases occur during an outbreak. Person-time incidence rate is the frequency at which new health related cases start to occur in the population. Crude rate is an outcome calculated not including restrictions such as age or gender. |
: Incidence rate= new cases occurring/population at risk *10^z |
: Point-Prevalence rate= existing cases at point in time/total study population at point in time *10^z |
: Attack rate= new cases occurring during short time/population at risk at start of short time *10^z |
Measures of Association
A contingency table is when every entry of data is classified by variables. The independent variable is an exposure and the dependent variable is the health related event. Correlation coefficient measures the strength of association between two variables. The geometric mean compares to the arithmetic mean on a logarithmic scale. Standard deviation is used in epidemiological studies. |
|
