Experimental Design

rules for setting up graph (11 things)

• x and y should be labeled

• scale should take up at least 75%

• x and y should be continuous and evenly distributed scale with no breaks

• x and y dont need to start at 0

• DRY MIX

• only major guidelines should be labeled

• all grpahing should be in pencil

• ont need to label every number

• points should be plotted with a X

• use a straight edge

• do not extrapolate

types of graphs (4)

• line graphs

• scatter plots

• bar chart

• histogram

line graph

• used for continuous variable

• automatically flows from one to the next

multiple-line line graph

• has more than one line plotted on the same set of axis

• lines should be drawn so there is a key difference

• key should be included or lines should be labeled

multi-axis line graph

• has more than one y-axis with different labeles and scales

• used to compare independent variable to two dependent variable

• more than one line so needs to be labeled

scatter plot

• not joined with straight lines

• straight line is drawn through several points keeping 50% of points above and below

bar charts

• show the relationship between discontinous variables

• do not touch

• uniform width and uniform spacing between them

• shouldnt be shaded or colored in

histogram

• used to compare continous data arranged in sets

• the bars touch

• bars should be the same width

• shouldnt be colored in

• bars are labeled with minimum to maximum value

• data set with no value are still included

when asked to interpret information from a graph

• provide general trend in data
  - direct relationship
  - inverse relationship
  - no relationship
  - linear, exponential or plateau

• quote data from graph
  - highest of lowest value, something that stands up

• manipulate data
  - provide range, percent change

significant figures rules

• all non-zero numbers are significant

• zeros between non-zero numbers are significant

• leading zeros are not significant (numbers to our left)

• trailing zeros in a whole number with no decimals arent signifcant

• trailing right of a decimals are significant

• an exact number has an infinte number of significant figures

calculating with signifcant figures

• gives us the precision of the measurements

• when we calculate from these measurements the answer can not be more precise than measurements used to determine it