Biostats Modules 3 & 4

Question 1

Parametric Test

Answer 1

• more powerful (more likely to detect a difference that truly exists)
• Assumptions:
• random independent samples
• interval/ratio level or measurement
• no outliers
• normally distributed
• homogeneity of variance
• less likely to make a type II error

Question 2

Type II error

Answer 2

-fail to reject null hypothesis when it was in fact false

Question 3

Type I error

Answer 3

-reject null hypothesis but is was in fact true

Question 4

Non parametric tests

Answer 4

• more conservative
• less statistical power
• more likely than parametric to produce type II error

Question 5

Benefits of using graphs

Answer 5

• summarizing data
• detects trends over time
• display patterns in large amounts of data
• analyze relationships among variables
• use graphs as an alternative to tables with many entries; do not duplicate data in graphs and tables

Question 6

Graphing Rules/Conventions

Answer 6

• the x-axis is formally known as the abscissa and the y-axis is the ordinate
• intersection of x and y is zero (not identified on graph)
• axes go from low to high
• x-axis portrays score values, categories, time
• y-axis depicts quantities like frequency, proportion, percent, cumulative frequency, cumulative proportion and cumulative percent

Question 7

Bar graphs/charts

Answer 7

• a type of column graph and are used with qualitative (non-numeric – nominal or ordinal) data along the x-axis.
• break categorical data down by groups on the x-axis, showing the frequency, relative frequency or percent for each group on the y-axis

Question 8

Bar graphs benefits

Answer 8

-preferred vehicle for communicating absolute information because they combine a picture with numbers, with each reinforcing the other

Question 9

Bar graph drawbacks

Answer 9

-can quickly become complicated once you move beyond a simple one-part bar

Question 10

Histogram

Answer 10

• used to show frequency distribution of data measured on interval/ratio level
• they show whats happening at a specific point in time rather than changes that occur over a long period
• “snapshots”
• can calculate skewness
• can help u gauge what needs further investigation
• not a stand alone tool for determining cause/effect

Question 11

Pie Graphs/charts

Answer 11

• alternative form of frequency chart (besides bar and histo)
• best used for nominal or ordinal data
• uses percentages to compare info
• drawbacks include, important info missing (such as total number of individuals)
• categories arent broken down sufficiently (i.e. “other”)
• not useful when values are similar

Question 12

Scatter Graph/scatterplot

Answer 12

• graphs designed to illustrate statistical correlation (NOT cause and effect) between both variables
• referred to as an XY values
• origin point (point where x and y axes intersect) is always (0, 0)
• benefit, researchers can confirm existence of possible relationship before investing time/money in causation study
• used alone, cannot show causation

Question 13

Line Graph

Answer 13

• similar to scatter graph except points are plotted in a sequence
• depict continuous data over time and are developed by joining a series of points with a line to show how a variable changes over time
• developed by joining a series of points with a line to show how a variables changes over time
• showing trends over time and how variables vary together