Standard Error (SEM)

Page 1

Perform data analysis following the data

Need to know: is there a statistically significant difference?

Is data truly different because of the independent variable, or are they diferent simply because of random chance

Mean: average value of the data

Light % - 74.9%

Dark % - 25.1%

If enough measurements are made, data in biology often follows a normal distribution ("Bell curve")

Standard Deviation: variance or spread of a sample

If SD close to 0, values are very close to mean value

If SD high, values are more spread out

Formula to calculate SD: on the AP exam list, but you will not have to use it for calculations

X with the line over it = mean

In normal distribution, 68% of data will fall within +/- 1 SD, and 95% will fall within +/- 2 SD's

Page 2

Standard error of the mean (SEM): a statistical test that puts an error range around our data means

SEM bars: can be added to graphs

+/- SEM 1 = 67% confidence that range includes true population mean

+/- SEM 2 = 95% confidence that range includes true population mean

We can use the SE bars to analyze the data

Check to see if the error bars overlap

If error bars do not overlap: the two groups ARE significantly different from each other; experimental hypothesis supported, null rejected

If error bars do overlap: the two groups are NOT significantly different from each other; experimental hypothesis is not supported, null is supported