Table of Contents

• Section 1. Frank's Story
• 1: The story
• 2: Biological interlude
• Section 2. How the meningitis population grows
• 3: How many meningococci can kill you?
• 4: Build the basic growth equation
• 5: Draw exponential growth
• Section 3. So why isn't Frank dead already?
• 6: How long has poor Frank got?
• 7: All we need now is data
• 8: Looking at data
• 9: Log transform straightens out exponential growth
• 10: Do ALL curves get "straightened" by taking the log?
• 11: Phases of bacterial growth
• Section 4. How can we tell when the population is growing exponentially?
• 12: A test for exponential growth
• 13: And the answer is...
• 14: No way its growing that fast!
• 15: Some more practice
• 16: In search of ... the exact doubling time
• 17: Finding doubling time with messy numbers
• 18: Logs = Exponent-busters
• 19: Let's bust some exponents
• 20: Let's bust some exponents
• 21: Extended problem #1: Exponential growth ends (with a whimper)
• Section 5. Back to Frank
• 22: Extended Problem #2: How long DOES poor Frank have?

Do ALL curves get "straightened" by taking the log?

Try it yourself and find out...

So, although log transformation is powerful, it is not ALL-powerful. The only curves above that get straightened are the ones generated by functions that include "some number to the power of the x variable". Why is that? Recall:

ANY population that grows or shrinks by a constant factor will appear as a straight line when log-transformed.

And that's what doubling does -- it multiplies the population by a constant factor (2, that is) over a constant period of time. But other mathematical functions, such as the ones above, involve some sort of change in which the factor of change is NOT constant. Let's take the parabola, for example: y = x²

x	1	2	3	4	5	6
y	1	4	9	16	25	36
factor		4	2.25	1.78	1.56	1.44

The factor by which y is increasing starts at 4, then drops to 2.25, then 1.78, and so on. Although we might think that a parabola grows "pretty fast", it is in fact "slow" compared to exponential growth that starts with the same factor.

As a general guideline:

• if growth factors are increasing, then growth is proceeding faster than exponential, and that part of the log-transformed curve will have a relatively steep slope.
• if growth factors are decreasing, then growth is proceeding slower than exponential, and that part of the log-transformed curve will have a relatively gentle slope.
• Only constant growth factors (exponential growth) leads to a straight-line when the graph is log-transformed.