Using the TI 83/84 Summary Statistics • Press STAT, ENTER (for EDIT) • If you do not see L1, scroll left. If you still don’t see it, press INSERT, 2nd, L1 (this is the 1 key) • If there are old data under L1: – Press the up arrow, then CLEAR, ENTER – DON’T use DELETE, ENTER. This causes L1 to disappear instead. Use INSERT, 2nd, L1 to get it back. • Enter data values in L1 one at a time, pressing ENTER after each. – If you make an error, use the up or down arrows to highlight the error, then enter the correct value. Use the arrows to get to the bottom of the list for the next value, if necessary. – Be sure to press ENTER after the last data value. Summary Statistics, Continued • Press STAT, Right Arrow (for CALC), ENTER • Press ENTER (for 1-Var Stats) • Press ENTER again • Read results – The Standard Deviation is labeled Sx Normal Probablilities • Find P(90 < x < 105) if x follows the normal model with mean 100 and standard deviation 15: • P(90 < x < 105) x x = normalcdf( 90 , 105 , 100 , 15) = .378 1 2 Normal Quantiles • We must find a so that P(x < a) = 2% when x has a normal distribution with a mean of 100 and a standard deviation of 15. • With the TI 83/84: a = invNorm( .02, 100 , 15) = 69.2 x Calculating r and Regression Coefficients • The first time you do this: – Press 2nd, CATALOG (above 0) – Scroll down to DiagnosticOn – Press ENTER, ENTER – Read “Done” – Your calculator will remember this setting even when turned off Calculating r and Regression Coefficients Continued • Press STAT, ENTER • If there are old values in L1: – Highlight L1, press CLEAR, then ENTER • If there are old values in L2: – Highlight L2, press CLEAR, then ENTER • Enter predictor (x) values in L1 • Enter response (y) values in L2 – Pairs must line up • Press STAT, > (to CALC) • Scroll down to LinReg(ax+b), press ENTER, ENTER • Read a, b, r and r2 Binomial Probabilities • To find binomial probability distribution values – probabilities for a particular number of successes, say 3 successes in 5 trials with the chance of success = .9: – – – – Let n = 5, p = .9, and x = 3 (Note the order: n, p, x) Press 2nd, DISTR [VARS] Scroll down to 0 and press ENTER, or just press 0 “binompdf(“ appears. Press 5, .9, 3 ) ENTER (Note that the commas are required, and note the order: n, p, x) – .0729 appears – If the values for x = 2, 3, and 4 are all needed • • • • Press 2nd, DISTR [VARS] Scroll down to 0 and press ENTER, or just press 0 “binompdf(“ appears. Press 5, .9, {2, 3, 4} ) ENTER All 3 values appear – In general, press 2nd, DISTR, 0, then enter n, p, x Binomial Probabilities Continued • To find the binomial probability for a range of successes, say 3 or fewer successes in 5 trials with the chance of success = .9 : – – – – Let n = 5, p = .9, and x = 3 Press 2nd, DISTR [VARS] Scroll down to A and press ENTER “binomcdf(“ appears. Press 5, .9, 3 ) ENTER (Note that the commas are required) • .0815 appears – If the probability for fewer than 3 successes is needed: • Enter binomcdf(5, .9, 2 ); read .00856 – If the probability for more than 3 successes is needed: • Enter 1 - binomcdf(5, .9, 3 ); read .919 Confidence Intervals for Proportions • Press STAT and use the cursor to highlight TESTS. • Scroll down to A: 1-PropZInt… and press ENTER. • The TI remembers your previous problem and shows the entries for it. You will overwrite these with the new entries. After x: enter the number of successes in the sample. • After n: enter the sample size. • After C-Level: enter the confidence level as a decimal fraction. • When the cursor blinks on Calculate, press ENTER again. Hypothesis Test for Proportions • Of 4276 households sampled, 4019 had telephones. Test, at the 1% level, the claim that the percentage of households with telephones is now greater than 93%. = .01 Claim: p > .93 H0: p < .93 .93 After p0 : enter .35 (the null hypothesis value) After x : enter 4019 (number of successes in the sample After n : enter 4276 (or p = .93) (the sample size) HA: p > .93 After prop highlight > p0 Press STAT, TESTS, 1-PropZTest, Enter (this is H A ) Highlight CALCULATE Press ENTER 11 Hypothesis Test for Proportions, Continued • Of 4276 households sampled, 4019 had telephones. Test the claim that the percentage of households with telephones is greater than the 93%. H0: p < .93 (or p = .93) HA: p > .93 Read 1-PropZTest .93 prop > .35 Test Performed HA 2.54 z-score z 80.87315928 p-Value p .00560 0 Sample Estimate .940 for p pˆ .9398971001 n 4276 Sample Size 12 Hypothesis Test for Two Proportions Example: For the racial profiling sample data, use a 0.05 significance level to test the claim that the proportion of black drivers stopped by the police is greater than the proportion of white drivers who are stopped. We have that: nB= 200 xB = 24 nW = 1400 xW = 147 And, H0: pB ≤ pW HA: pB > pW On the TI-83/84: Press STAT, arrow to TESTS and scroll down to 2-PropZTest. Press Enter. Make these entries: Highlight Calculate and press ENTER 13 Hypothesis Test for Two Proportions, Continued Your TI screen should look like this: This is HA This is the p- value. This is p If you press the down arrow twice, the TI will tell you the sample sizes, as a check. 14 Confidence Interval for a Mean n = 106 y = 98.20o s = 0.62o = 0.05 15 Hypothesis Test for a Mean Claim : 1017 H 0 : 1017 H1 : 1017 TI 83 / 84 entries: STAT TESTS 2 Stats Enter 0 : 1017 x : 1040 sx : 207 n : 20 : 0 Enter TI 83 / 84 entries (continued): Highlight Calculate Enter Read >1017 ( H1 ) t .4969... p .312787... x 1040 sx 207 n 20 16 Hypothesis Test for Two Means Press STAT and arrow over to TESTS. Scroll down to 2-SampTTest and press ENTER. Highlight HA as μ1: ≠ μ2 Your screen should look like this: Highlight STATS and press ENTER Set μ0: 0 Enter McGwire’s stats for sample 1 McGwire n1 x1 s1 70 418.5 45.5 Enter Bonds’s stats for sample 2 Bonds n2 x2 s2 73 403.7 30.6 Select Pooled: No Select Calculate Press ENTER Hypothesis Test for Two Means, Continued Your screen should look like this: HA Test Statistic P-Value Scrolling down reveals some to the statistics you entered earlier. Because the P-value is less than .05, we reject The Null Hypothesis The sample data support the claim that there is a difference between the mean home run distances of Mark McGwire and Barry Bonds. Confidence Interval for Two Means Using the sample data given in the preceding example, construct a 95%confidence interval estimate of the difference between the mean home run distances of Mark McGwire and Barry Bonds. We need E so that (x1 – x2) – E < (µ1 – µ2) < (x1 – x2) + E As before, Press STAT, go to TESTS, and now scroll down to 2-SampTInt and press ENTER. The TI remembers the entries from the last time 2-SampTTest or 2-SampTInt was used. Since these are the same, we need only scroll down to C-Level to make sure it is set to .95 (NOT 95%!). Pooled is always No for our work. Highlight CALCULATE and press ENTER. Matched Pairs Test TI-83/84 users may test paired sample hypotheses from raw data as follows: Enter first half of each STAT pair in L2 EDIT Enter second half of each Highlight L1 pair in L3 CLEAR 2nd QUIT ENTER 2nd L2 – 2nd L3 STO> 2nd L1 Highlight L2 CLEAR STAT ENTER TESTS Highlight L3 2: T-Test CLEAR Highlight Data ENTER ENTER Proceed as for one-sample t - Test Matched Pairs Test, Continued Claim: there is a difference between the actual low temperatures and the low temperatures that were forecast five days earlier That is, μd ≠ 0 H 0: d = 0 H 1: d 0 Matched Pairs Confidence Interval Follow the instructions given above for entering the 2 samples into L2 and L3, then storing the difference in L1. We now need only compute a one-sample confidence interval using STAT TESTS TInterval ENTER Data List: L1 Freq: 1 C-Level: .95 Calculate ENTER Goodness-of-Fit Test 1. STAT, EDIT, CLEAR L1, L2 and L3 2. Enter observed counts in L2. 3. Enter expected counts in L3. 4. 2nd Quit Store (Obs-Exp) 2 in L1 Exp 5. STAT, CALC, 1-Var Stats, ENTER. Write down value of Σx 6. 2nd, DISTR, 8:X2cdf, ENTER, Σx, 1000, k-1) Read P-Value Test for Homogeneity Claim: There is no difference in the distribution of the continent of origin for student and staff cars. H0: There is no difference in the distribution of the continent of origin for student and staff cars. HA: There is a difference in the distribution of the continent of origin for student and staff cars. Set α = .05