regressions....THIS IS OUR LAST SHOT

Question 1

turning off scatterplot
go t oy= screen, arrow up onto plot highlighted at top of screen press enter to turn off

Question 2

line of best fit taken from linear regression equation (y= ax+b)

Question 3

(mean of x, mean of y) will ALWAYS be on the line of best fit.

Answer 3

when asked for proof, put mean of y at y= and basically solve/prove the answer correct

Question 4

if the correlation coefficient is negative, the line of best fit will have a negative slope

Question 5

the correlation coefficient will always be the one closest to 1 or -1

Answer 5

The one with the weakest correlation will be the one closest to zero

Question 6

if a set of data points do not appear to form a linear relationship, then different nonlinear functions can be constructed from the data using the regression feature of the graphing calculator (quad, cubic, exp, power, sinusodial- radian mode!!)

Question 7

Stat, edit, plug in data, stat–> calc
#5- quadratic
#5- cubic
#0- exponent
#A–> power

Question 8

be very careful with how data is measured ex. if 2400 but measured by 00’s, you work with 24

Question 9

for equations like cubic that are wayy too complicated to do all at once put y= ansewer in y1 and the equation into y2 and find the intersection

Question 10

if there’s a decimal number in an “initial amount” and you’re asked why it’s not an integer, it’s because the model isn’t an exact fit ex. 10.18 vs 10

Question 11

for exponent questions, use calculator as much as possible

Answer 11

be extremely careful for data that says for ex. let first x= 0 because that affects the rest of the data ex. 1993=0, 1995= 2, 1999= 6, etc etc

Question 12

Solving EXPONENT EXAMPLE
100= 17.82(1.1319) to x power
- divide both by 17. 82
-the x power turns into a log on the 100 side and a log next to 1.1319 and an x beside it so it looks like: log 5.62366438= log1.319x , then divide both sides by log 1.1319 (and make sure to include the log)

Answer 12

“in what year will there be 100 mil subscribers?”

Question 13

when asked if the data either appears exponential or logarithmic, look at the shape of the plot (if it went up but flattening, it’s log, if it’s j shaped it’s exponent)

Question 14

when finding r= for 3 different equations (ex. linear, log, and exp) usually it’s either the line is closest to -1 0 or 1 so pick the correlation that seems to match/is the highest or lowest

Question 15

solving LOG EXAMPLE (lnx) **INCLUDES E
103 = -2.3 + 31.2 lnx
- add 2.3 and divide 31.2 to 103 so you just have lnx by itself which becomes lnex
move e over to side of 100 (leave x on right side) so it looks like
e(102.3/31.2)= x

Question 16

fit an exponential function when modeling population growth for a similar situation

Question 17

power functions often go with scientific measurements!!

Answer 17

• fit a power function to area or volume measurements
• fit a power function when scatter plot indicates that both coordinate axes are asymptotes
• fit a power function rather than an exponential function if (0,0) is a possible data point

Question 18

SINUSODIAL GRAPH- must be in RADIAN MODE

Answer 18

a= amplitude
b= frequency where b>0
2pi/b = period
c/b horizontal shift
d- vertical shift (a+d is max, a-d is min)

Question 19

c/b sinusoidal graph

Answer 19

horizontal shift (if c<0 to the right and if c>0 to the left)

Question 20

period

Answer 20

horizontal distance between 2 maximum or 2 minimum points

Question 21

**No correlation coefficient with sinusoidal graphs