Mr.

Page
Room 154
AP Statistics
Calculator Instructions
This is a “little” cheatsheat for you to use to review how to do certain critical operations on the
calculator. The instructions are for a TI-83, because to my knowledge that is what almost
everyone is using. If you would like one for the TI-89, let me know. I’m not very good with it,
though. Operations underlined are the actual keys. Operations in italics are NOT critical but
might be useful. They probably aren’t worth learning if you have yet to do so. NEVER EVER
WRITE CALCULATOR DIRECTIONS AS PART OF AN ANSWER.

Overview
Entering data STAT. EDIT.

Calculating information about a set of STAT. CALC

data

Calculating probabilities 2nd. DISTR.

Plotting data 2nd. STATPLOT.

Generating data according to a MATH. PRB.

probability

Inference (tests/confidence intervals) STAT. TESTS.

Summarizing Data
Summary statistics of a set of data STAT. CALC. 1-VAR STATS. List #
-gives 5 # summary, mean, standard deviation
(remember to read s if the data is a sample (99%
of the time), σ only when it’s a population)

Histograms, boxplots STATPLOT

- remember to set the window, or you can use
ZOOM. ZOOMSTAT. to focus the window.
- for histograms, use Xscl under WINDOW to
set the width of each bar
- use trace to identify values quickly
- use the box plot with dots (on the left) for a
modified plot (usually what will be used)
- if you’re given values and counts (e.g., there
were 10 people with 5’s, 8 people with 4’s, 9
people with 3’s, and 3 people with 2’s), rather
than raw data, you can enter each value
(5,4,3,2) into L1, and you can enter the counts in
L2 (10,8,9,3), and then set Freq to equal L2

Normal Distribution
Area under a normal curve DISTR. Normalcdf.
- enter normalcdf (lower value, upper value,
mean, s.d.)
- enter normalcdf (s.d.s to the left, s.d.s to the
right) if your values are normalized (z-scores),
i.e., mean = 0 and s.d. = 1
- if you only want the area to the right of a
certain value, enter a BIG number for the max
value). Vice-versa for area to the left of a
certain value and min.
Mr. Page
Room 154
AP Statistics
- use DISTR. DRAW. Shade in order to see the
normal curves (t curve and chi-square as well)

To find out the value and/or number DISTR. invNorm

of standard deviations that capture a - enter invNorm (area under curve to the left-as
certain percentage of the density curve a decimal, mean, s.d.)
- for # of s.d.s, enter invNorm (area), and you
will get how many s.d.s to the right and left

Normal probability plot STATPLOT.

-then enter the straight line on the lower right
option. If the line’s linear, the data is normal. If
it’s curved or shaped otherwise, it’s not normal.
If a point is way off of the line, it’s an outlier.
- the data does NOT need to be sorted ahead of
time.
Linear Regression
Scatterplot - enter x-values and y-values in separate lists (use
L1 and L2, respectively, to keep it simple)
STATPLOT
-use scatterplot box, and use ZOOMSTAT to
focus on the data
-ALWAYS ALWAYS ALWAYS look at this
before doing any regression calculations

Linear Regression STAT. CALC. LinReg(a+bx) (L1,L2)

-output should include intercept, slope, r2, and r.
-if you don’t see r2 and r, then go to CATALOG,
and scroll down to DIAGNOSTIC ON, and hit
enter.

Residuals -after calculating a regression, you can view

residual values by going to LIST and scrolling
down to RESID
-you can do a residual plot by doing a scatterplot
with your original x-value as x, and RESID as y
-you can view statistics on the residuals by
entering 1 var stats. RESID.

Non-linear relationships -do these when the scatterplot appears curved

(exponential, power) -take log of L1 (x) first, and then do a regression
of log x on y to test for an exponential
relationship
-take log of L2 (y) as well and regress log x on
log y to test for a power relationship

Random Variables
To generate random integers (useful for MATH.PRB.RANDINT (min#,max#, # of #s you want)
when you have to do a simulation on
a free response question)

To generate a random # between 0 and 1 MATH.PRB.RAND

To generate random numbers from a MATH.PRB.RANDNORM (mean, s.d., # of #s to

normal distribution generate)
Mr. Page
Room 154
AP Statistics
To generate a random binomial number MATH. RANDBIN. (enter 1 for success/2 for failure,
probability of success, # of #s to generate)

Binomial Distribution
Probability of a count (e.g., if I roll a DISTR.BINOMPDF. (# of trials, probability of success,
die 10 times, what are the odds I get a # of successes)
6 exactly 3 times?) -binompdf (10, 0.166667, 3)

Probability of a count or less (e.g., if I DISTR.BINOMCDF. (# of trials, probability of success,

flip a coin 12 times, what are the odds max # of successes)
I get heads 4 or fewer times?) -binomcdf (12, 0.5, 4)
-from the previous example, if you want to know
“what are the odds I get heads 5 or more times,”
the easiest way is to calculate 1 – binomcdf (12,
0.5, 4)

Inference
NOTE: I’m lazy, and I still prefer to do these by hand unless they’re very time consuming, e.g.,
chi-squared, because it’s the only way I can be confident of my answer. If you have time, you
should do both!!!

Confidence interval for known s.d. STAT. TESTS. Z-INTERVAL.

-you can enter the raw data in the list, or simply
enter the key statistics (mean, s.d., n, conf. level)
-be sure you get whether it’s a 1 or 2 way test
-always write out the formulas on your answer

Hypothesis test for known s.d. STAT.TESTS.Z-TEST.

-very similar to above, but you need Ho here

2 Sample C.I./Hypothesis test for STAT.TESTS.2-SAMP Z-INTERVAL/TEST

known s.d. (very rare!)

C.I./Hypothesis test for unknown s.d. STAT.TESTS.T-INTERVAL/T-TEST

2 Sample C.I./Hypothesis test for STAT.TESTS.2-SAMP T-INTERVAL/TEST

unknown s.d.s

Confidence interval/hypothesis test for STAT.TESTS.1 PROP Z-INTERVAL/TEST

proportions

Confidence interval/hypothesis test for STAT.TESTS.2 PROP Z-INTERVAL/TEST

2 proportions

Chi Square test for goodness of fit, STAT.TESTS.X2 TEST

association, independence, etc. -note: it’s best to enter the data in to a matrix
first. To do so hit MATRX.EDIT, and enter the
amount of rows and then columns

Testing for the slope of line STAT.TESTS.LinRegTTest

-note: you need to have raw data in a list first
-be sure you know what you’re checking for
about the slope of the line
Mr. Page
Room 154
AP Statistics
To restore a deleted list (if the list is L1 through L6), just go to
the position that the list should be in, hit 2nd insert and then hit 2nd and the list number on the calculator.
For example, if List 5 was inadvertently deleted, highlight L6 on the calculator, press 2nd insert, and a open
space will present itself, then press 2nd L5 (L5 is above number 5 on calculator) and List 5 should be
restored.

Dolores Angelino
Bishop Eustace Prep
Pennsauken, N.J.