Finding Random Numbers on a TI 83, 83+, 84+ Calculator

Math PRB (probability menu)

1:rand generates a random number between 0 and 1 5:randInt ( m , n , k ) generates k random integers between the integers m and n inclusive (m < n). Put in the appropriate numbers for n, m, and k.
Example: randInt(1,6,2) randomly selects 2 integers between 1 and 6 and returns them in the form of a list between brackets {5 1}. RandInt{1, 6, 2} therefore simulates the act of rolling two dice.

For later in the term: 6:randNorm ( , , k ) generates k random values from a Normal probability distribution with mean and standard deviation . Use appropriate values for k

7:randBin ( n , p , k ) generates k random values from a Binomial probability distribution with n trials and p = probability of success.

TI 83, 83+, 84+: Entering Data into the statistics list editor
STAT EDIT Put data into list L1, press ENTER after each data value 2nd QUIT to exit stat list editor after you have entered data, checked it and corrected errors.

TI 83, 83+, 84+: Sorting Data

2nd QUIT to exit the statistics list editor STAT EDIT 2 for SORTA sort in ascending order ( 2n d L1 ) The calculator responds DONE STAT EDIT to view the sorted data now stored in L1 2nd QUIT exits the statistics list editor,

FOR ALL TI GRAPHING CALCULATORS When you put data into the lists there are two methods to handle repeated values.
Example: Suppose your data is 1, 4, 4, 4, 5, 7, 8, 8.
2 ways to enter data Enter data in a single list L1 (list each item separately) 1 4 4 4 5 7 8 8 O R Use a data list and a frequency list L1 (data values) 1 4 5 7 8 L2 (frequencies) 1 3 1 1 2

What the data looks like:

For 1 variable statistics: For statistics plot:

TI 86: TI 83, 84:

One Var L1 1Var Stats L1 Xlist: L1 Freq: 1

TI 86: TI 83, 84:

One Var L1, L2 1Var Stats L1, L2 Xlist: L1 Freq : L2

TI 83, 83+, 84+: One Variable Statistics

STAT CALC 1 for 1 Var Stats : one variable summary statistics 2nd L1 ENTER. If data is in a different list indicate the appropriate listname instead of L1

Interpreting The Output For One Variable Statistics

X
x x2 sx mean sum of all data values squares each data value, then adds them up sample standard deviation population standard deviation number of data values minimum (smallest) data value value of first (lower) quartile median value of upper (third) quartile minimum (smallest) data value We call this for a population or a probability distribution; calculator always uses X

x
n MinX Q1 Med Q3 MaxX

The calculator doesnt know if your data is a sample or a population, or a probability distribution. You need to know whether to select the use sx or x

TI 83, 83+, 84+: Graphing Univariate (One Variable) Data

Histogram using the TI 83, 83+, 84+: 2nd STATPLOT 1 Highlight ON ; press ENTER To select Type: Highlight histogram icon To set Xlist: 2nd L1 ENTER To set Freq: 1 ENTER ZOOM 9 Calculator constructs the histogram, determining its own intervals. Be aware that the calculator may select "strange" intervals because it has no understanding of the context of the data. To specify your own histogram intervals, adjust the window: WINDOW XMin= lower boundary of first interval Xscl = interval width XMax = upper boundary of last interval press ENTER

Example: For intervals 1519, 2024, 2529, . . ., 6064, 6569: Xmin = 14.5 Xmax=69.5 Xscl=5

YMin = 0 Estimate YMax to be large enough to display the bar for the interval with the highest frequency. Select an appropriate value of YScl for the tick marks on the y-axis GRAPH Calculator constructs the histogram with user specified intervals

TRACE then use right and left cursors to move the cursor from bar to bar. The screen indicates the frequency (height) for the bar that the cursor is positioned on. Boxplot using the TI 83, 83+, 84+: 2nd STATPLOT 1 Highlight ON ; press ENTER To select Type: Highlight boxplot icon To set Xlist: 2nd L1 ENTER To set Freq: 1 ENTER ZOOM 9 to graph the boxplot TRACE then right and left cursors move the cursor between the five values defining the boxplot (min, Q1, median, Q3, max). The screen will indicate the appropriate numerical value. To turn off statistics plots: 2nd STATPLOT 4 (for Plots Off) will turn all statistics plots off. press ENTER

DIMENSION ERROR message

If you get a dimension error, you probably have one of two types of errors. One of these methods will usually eliminate a dimension error:

You may be trying to do something with two lists that are supposed to be the same length but for some reason are not. Check that you have the correct information in each list or that you are using the correct lists. Check to be sure that your list is not empty. You may be trying to graph an equation in your equation editor at the same time as you are trying to graph at STATPLOT. If you want to graph the equation, turn your statplots off. If you want to draw the statplot, go into the Y= equation editor and turn off or delete the functions that are stored in Y1, Y2, . . .

PROBABILITY DISTRIBUTIONS SUMMARY on the TI 83, 83+, 84+, 86, 89 TI-83 and 84, press 2nd DISTR TI-86 press 2nd MATH MORE; then press F2 or F3 for the STAT menu; press F2:DISTR TI-89 press APPS; Press 1: FlashApps; highlight Stats/List Editor press ENTER F5: Distr
DISCRETE PROB ABILITY DISTRIBUTIONS: function & input TI 83,4: binompdf(n,p,r) TI 86: bipdf(n,p,r) TI 89: binomial Pdf TI 83,4: binomcdf(n,p,r) TI 86: bicdf(n, p, r) TI 89: binomial Cdf pdf gives P(x = specified value) cdf gives P(X specified value) parameters Description n = number of trials Binomial probability P(X = r) of exactly r p = probability of success successes in n independent trials, with r = number of success probability of success p for a single trial. If r is omitted, gives a list of all probabilities from 0 to n n = number of trials Binomial cumulative probability P(X r) of r p = probability of success or fewer successes in n independent trials, r = number of success with probability of success p for a single trial. If r is omitted, gives a list of all cumulative probabilities from 0 to n p = probability of success Geometric probability P(X = n) that the first n = number of trials success occurs on the nth trial in a series of independent trials, with probability of success p for a single trial. p = probability of success Geometric cumulative probability P(X n) n = number of trials that the first success occurs on or before the nth trial in a series of independent trials, with probability of success p for a single trial. Poisson probability P(X = r) of exactly r = mean occurrences for a Poisson distribution with r = number of occurrences mean = mean Poisson cumulative probability P(X r) of r or r = number of occurrences fewer occurrences for Poisson distribution with mean

TI 83,4: geometpdf(p,n) TI 86:gepdf(p,n) TI 89:geometric Pdf TI 83,4: geometcdf(p,n) TI 86:gedf(p,n) TI 89: geometric Cdf TI 83,4: poissonpdf( ,r) TI 86: pspdf(,k) TI 89:Poisson Pdf TI 83,4: poissoncdf(,r) TI 86:psdf(,r) TI 89:Poisson cdf

CONTINUOUS PROB ABILITY DISTRIBUTIONS "cdf" functions find the probability as area under the curve, above x axis within a specified interval of x values. "pdf" functions find the height of the curve above the x axis at a single x value; the "pdf" functions do not find probabilities, but they enable the calculator to draw or graph the curves for the probability distributions. The inverse functions find the value of a percentile. The only inverse function on the TI-83 and TI-86 is the inverse normal. The TI-84 and 89 have expanded selections of inverse functions. function & input parameters Description TI 83,4: c = lower bound P(c < X < d) for a normal distribution with mean normalcdf (c, d, , ) d = upper bound and standard deviation . = mean To find P(X>c) use upper bound = 10 ^ 99. TI 86:nmcdf (c, d, , ) = standard deviation To find P(X < d) use lower bound = () 10^ 99, TI 89:Normal Cdf using () key to indicate a negative number. TI 83,4: invNorm(p, , ) p = percentile = area to the left Finds the value of x =c for which P(X < c) = p = mean for a normal distribution with mean and TI 86:invNm(p, , ) = standard deviation standard deviation . TI 89: 2:Inverse } 1: Inverse Normal TI 83,4: tcdf(a, b, df) a = lower bound Probability that a value lies between a and b for a b = upper bound Student's t distribution with the specified degrees df = degrees of freedom of freedom To find P(X > a), use upper bound = 10 ^ 99. To find P(X < b), use lower bound = () 10 ^ 99, using () key to indicate a negative number. a = lower bound Probability that a value lies between a and b for TI 83,4: 2 cdf(a, b, df) b = upper bound a 2 chi-square distribution with the specified TI 86:chicdf(a, b, df) df = degrees of freedom TI 89:Chi-square Cdf degrees of freedom

To find P(X > a), use upper bound = 10 ^ 99. To find P(X < b), use lower bound = 0.

Binomial, Geometric, Poisson Distributions on the TI 83, 83+, 84+, 86, 89: TI-83 and 84, press 2nd DISTR TI-86 press 2nd MATH MORE and then press F2 or F3 for the STAT menu item F2:DISTR (menu location of STAT may vary on different calculators) TI-89 press APPS; pPress 1: FlashApps; highlight Stats/List Editor press ENTER F5: Distr pdf stands for probability distribution function and gives the probability P(x = r) cdf stands for cumulative distribution function and gives the probability P(x r) Binomial Distribution TI 83, 84 P(x = r) P(x r) P(x < r) P(x > r) P(x r) binompdf(n,p,r) binomcdf(n,p,r) binomcdf(n,p,r1) 1 binomcdf(n,p,r) 1 binomcdf(n,p,r1) TI86 bipdf(n,p,r) bicdf(n,p,r) bicdf(n,p,r1) 1 bicdf(n,p,r) 1 bicdf(n,p,r1) TI-89 binomial pdf(n,p,r) binomial cdf(n,p,r) binomial cdf(n,p,r1) 1 binomial cdf(n,p,r) 1 binomial cdf(n,p,r1)

Geometric Distribution TI 83, 84 P(x = n) P(x n) P(x < n) P(x > n) P(x n) geometpdf(p,n) geometcdf(p,n) geometcdf(p,n1) 1 geometcdf(p,n) 1 geometcdf(p,n1)

TI86 geopdf(p,n) geocdf(p,n) geocdf(p,n1) 1 geocdf(p,n) 1 geocdf(p,n1)

TI-89 geometric pdf(p,n) geometric cdf(p,n) geometric cdf(p,n1) 1 geometric cdf(p,n) 1 geometric cdf(p,n1)

Poisson Distribution TI 83, 84 P(x = r) P(x r) P(x < r) P(x > r) P(x r) poissonpdf(mu,r) poissoncdf(mu,r) poissoncdf(mu,r1) 1 poissoncdf(mu,r) 1 poissoncdf(mu,r1)

TI86 pspdf(mu,r) pscdf(mu,r) pscdf(mu,r1) 1 pscdf(mu,r) 1 pscdf(mu,r1)

TI-89 poisson pdf(mu,r) poisson cdf(mu,r) poisson cdf(mu,r1) 1 poisson cdf(mu,r) 1 poisson cdf(mu,r1)

TESTS FUNCTIONS SUMMARY on the TI 83, 83+, 84+, 86, 89

TI-83: STAT }} TESTS TI-86: 2nd MATH MORE F2 or F3:STAT F1:TESTS

(menu location of STAT may vary on different calculators) TI-89: APPS 1: FlashApps Highlight Stats/List Editor and press ENTER 2nd F6: Tests
Hypothesis test for a single mean, population standard deviation known Hypothesis test for a single mean, population standard deviation unknown, underlying populations approximately normally distributed Hypothesis test of the equality of two population means, independent samples, population standard deviations known Hypothesis test of the equality of two population means, independent samples, population standard deviations unknown, underlying populations approximately normally distributed Hypothesis test of a single proportion Hypothesis test of the equality of two population proportions Confidence interval for a single mean, population standard deviation known Confidence interval for a single mean, population standard deviation unknown, underlying populations approximately normally distributed Confidence interval for the difference between two means, population standard deviations known Confidence interval for the difference between two means, population standard deviations unknown, underlying populations approximately normally distributed Confidence Interval for a single population proportion Confidence Interval for the difference between two population proportions Hypothesis test of independence for a contingency table stored in a matrix. Expected values are calculated and placed in a separate matrix Hypothesis test for Goodness of Fit Observed and expected data counts must be placed in lists Hypothesis test of the equality of two population standard deviations Hypothesis test of the signficance of the correlation coefficient in linear regression Hypothesis test of the equality of means of multiple populations using one way analysis of variance for sample data entered into lists Z-Test ZTest T-Test TTest 2-SampZTest Zsam2 2-SampTTest Tsam2 1-PropZTest ZPrp1 2-PropZTest ZPrp2 ZInterval ZInt1 TInterval TInt1 2-SampZ Int ZInt2 2-SampT Int TInt2 1-PropZInt ZInt1 2-PropZInt ZInt2 2 Test Chitst Chi2 2-way 2 GOF Test (TI-84 & 89 Only) 2-SampFTest Fsam2 LinRegTTest TLinR ANOVA

Linear Regression and Correlation: Drawing a Scatterplot on the TI 83, 83+, 84+ in a window sized to show the data
TI 83, 83+, 84+: 2nd STATPLOT 1 On Off Type Highlight the scatterplot icon and press enter Xlist: list with x variable Ylist: list with y variable Mark: select the mark you would like to use for the data points ZOOM 9:ZoomStat

You can use TRACE and the right and left cursor arrow keys to jump between the data points and show their (x,y) values
To show the linear regression line, type the equation of the line into function Y1 in the Y= function editor. TI 83, 83+, 84+: Y= Enter the equation into Y1 as Y1 = a + bX, using the appropriate values for a and b in the equation. You can enter X using the X,t,,n key

Setting up your TI 83 & 84: 2nd catalog (above 0 key) D(above x 1 key) scroll down to Diagnostics On With the cursor pointing to Diagnostics on press Enter When Diagnostics On is pasted to your home screen, press Enter again You should only need to do this once. After you have turned Diagnostics ON it should stay on unless you turn it off, reset your memory, or your batteries and the backup battery both go dead.

Linear Regression Using the TI 83, 83+, 84+: Linear Regression t test
TI 83, 83+, 84+: STAT TESTS Xlist: list with x variable Ylist: list with y variable Freq: 1 & : 0 <0 >0 RegEQ: leave blank Calculate ENTER E:LinRegTTest

OUTPUT of Linear Regression t test LinRegTTest y = a + bx 0 and 0 t = test statistic p = p-value df = n - 2 a = value of y-intercept b = value of slope s = standard error r2 = value of coefficient of determination r = value of correlation coefficient

The linear regression equation is stored in the graphing equation editor as function Y1 on TI-83, 83+ & 84+. If you select this function to be on when you graph the scatterplot of the data, the graph will also show the least squares linear regression line on the scatterplot of the data If you have trouble entering the equation name into the test window, you can type the equation into Y1 manually after you perform the test (see scatterplot instructions on previous page)

Linear Regression and Correlation: Identifying Outliers in Bivariate Data graphically using the TI calculators
Perform the linear regression using the Linear Regression T Test Function Write down the value of the Standard Error s, the slope b and the y-intercept a Go to the Y= equation editor. Type the linear regression equation into equation Y1: Y1 = a + bX, using the appropriate values for a and b in the equation. You can enter X using the X,t,,n key Enter two new equations: Y2 = a + bX 2 * s and Y3 = a + bX + 2 * s using the appropriate values of a, b, and the standard error s in the equations Draw the graph: press ZOOM 9:ZoomStat An outlier is any point for which the data value is more than two standard errors away from the linear regression line. Any points that lie above the top line or below the bottom line are outliers, being more than two standard errors away from the regression line. Any points that lie between the lines are NOT outliers. Use TRACE and the right and left arrow keys to identify which are the points that are the outliers and to find the x and y values for these data points.

Test of Independence for TI-83, 83+, 84+

1. Access MATRIX menu to input data to matrix A: TI83 : MATRIX EDIT ENTER nd MATRIX TI83+ & 84+ : 2 EDIT ENTER 2. Enter matrix dimensions: Type: # rows ENTER

# columns

ENTER

3. Enter table values going across rows Press ENTER after each value. When done, press EXIT. 4. Run program to do Chi Test

2
-TEST

TI83, 83+ & 84+ STAT TESTS C: Make sure Observed says A and Expected says B. Press CALC or Calculate

You will see the test statistic and p-value on the screen. 5. To see the expected values matrix TI83, 83+ & 84+ MATRIX 2

ENTER

Warning about Chi Square Test for Independence

When you use your calculator for a chi square test for independence you should ALWAYS use CALC (calculate) and you should not use DRAW. You should draw your graph yourself, using your knowledge of that this test is always a right tailed test. There is a bug in the calculator program for chi test using DRAW. Sometimes it shades the graph in the wrong direction. A chi test for independence is always a one tailed test to the right (because you are testing whether the chi square test statistic is a large number, indicating large differences between observed and expected).

If you have the calculator draw, it may shade to the right correctly, or it may incorrectly shade to the left instead. The direction of the shading appears to be related to the direction of the last z, t, or proportion hypothesis test that was done on the calculator. If the last test of a mean or a proportion was two tailed or right tailed, it seems to shade to the right, correctly. If the last test of a mean or a proportion was left tailed, it incorrectly shades left instead of right.