Stats Pad User Manual
Contents
1. u 3 Gar POE CENT c Dial Bin gt add C9 C4 CS C8 ENTE 07000 a z WEN NEHOW O 510361 522496 T eee TIT SUM v mI 60 0042 40 0633 57168 7 yy BEE 586977 74286 528197 amp 47 HEIDE ETT a zn E TS JE N pos Um 45 Proportions Tests Single Proportion Test A Single Proportion test will determine if a sample set of data comes from a population with the hypothesized proportion Set the hypothesized proportion as a decimal between 0 0 and 1 0 Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha value for the test Enter the number of observed successes as X Enter the total sample size as N Tap the calculate button Single Proportion Test For One Proportion Uses normal apperosmabon X Number of successes observed N Total Sample sie Hypothesized Proportion X Nut Hypothesis o EED E e Aha 05 Cak uste 46 Two Proportions Test A Two Proportions Test will determine if a two samples of data come from a populations with the hypothesized difference in proportions Set the hypothesized difference in proportions as a decimal between 1 0 and 1 0 Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha value for the test Enter the number of observed successes for each sample as X Enter the total sample size for each sample as N2. C1 c2 C3 I C4 C5 C8 lt with Replication ENS Jh ee R2 1lSMAL MEDIUM ILARGE R3 HOT 106322 110077 115 494 RA a ae pos I RS HOT 101 451 106812 138 298 J R6 WARMOS 14678 109 97 lessa RT TUM E Ra WARM 109965 113306 R9 C oes ses mesos nass 1 Rit OO 38 242 112329 10959 1 1 EN Select the range of cells containing the data One factor should be grouped by columns e other grouped by rows Rows per sample 3 Alpha 06 First rowcolunm contams labeis Cx Calculate 13 Kruskal Wallis ANOVA A Kruskal Wallis One Way ANOVA tests if multiple samples came from populations with the identical medians This test is typically performed when one factor is set to multiple levels to determine if that factor affects the median of the distribution Use this test when a standard One Way ANOVA is desired but the data does not meet the criteria of normality Stats Pad expects each set of sample data to be in adjacent columns In the example below we are testing if color affects our data measurement Select the region with the data If Labels are included as in this example set labels to ON Set Data grouped by Column if the data is in columns as in this example Set the desired alpha level and tap Calculate Cu Kruskal Wallis ANOVA Example Une mede Ld Au EE ae a EREMO Kruskal Wallis One Way A
3. Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button MCN AC 6 o oo ooo o For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button bt ch band 5 ee T RAT AI 4 9 9 _ 49 9 gt 9 0 o 9 gt gt o oe o o oeoo oeo 4 9 9999 ce oe 9 55
4. 51 0361 52 24968 R4 928 105 54 7966 56 1951 178 0513 RS 1302395 45 6948 47 4213 5819 ne B0063 440 0623 62 7666 186 0877 4 75 7168 LL A RIS 557092 521486 822828 625367 mm ET RE Tem M a TEE TT ER 11 Random Block ANOVA A Random Block ANOVA also known as a Two Way ANOVA without Replication tests if multiple samples came from populations with the identical means where two factors are vary There is only one data point per combination of settings for the two factors see the example below This test is typically performed when two factors are set to multiple levels to determine if either factor affects the mean of the distribution Stats Pad expects each set of sample data to be in adjacent rows and columns In the example below we are testing if color and Alpha setting affects our data measurement Select the region with the data If Labels are included as in this example set labels to ON Set the desired alpha level and tap Calculate Example Uno mede Ld T G1 I cz cs I c4 cs ce L 1 MEN CEREREM E Gu Random Block ANOVA Complete Randomized Blocking ANOVA Two F actor ANOVA without Replication Select the range of cells containing the data The main factor should be grouped by columns the blocking factor grouped by rows R2 f RED jBLUE GREEN YELLOW ee Rss 2228 _ s4 5971_ 51 0
5. Create Power Curve 48 Random Numbers Normal T Distribution Uniform Chi Square F Distribution Exponential The Random Numbers section provides a method of filling up a selected range of cells with random data following a selected distribution Currently the normal T uniform chi square F and exponential distributions can be used Select the range of cells to be fill with random data Choose the distribution to be used Normal for this example Fill in any required parameters for that distribution Tap the Generate button and the range will be filled with random data j ane We Normal Random i Normal Random Numbers FR selected range with random R4 Populaton Mean 250 m6 EE 0l oel A J pn EE EE 0l P 1 EEE er men ame Rit mu PD Pj P ee SS a a O SS SOHN 1 amp RIS R16 R17 EE l ol EN oL oc i RS 49 Regression Simple amp Multiple Simple and Multiple regression will calculate a least squares fit model that predicts Y values based on input X values This test will also provide all of the statistics necessary to determine if the model is valid and significant The following settings are available Force Y Intercept to O It is advised that this be left to OFF unless the user is very familiar with Regression Through the Origin and its potential drawbacks Confidence Level This will be used t
6. Pad is targeted to students taking a class in statistics or who have already taken a class who have a textbook that will explain how to interpret the results The output is designed to align very closely with how statistics textbooks approach analysis As this is the first revision of a Stats Pad manual focus has been placed on how to set up the data and run the test This version of the manual does not cover interpreting the output Future versions of the manual will add output interpretation 10 ANOVAS One Way ANOVA A One Way ANOVA tests if multiple samples came from populations with the identical means This test is typically performed when one factor is set to multiple levels to determine if that factor affects the mean of the distribution Stats Pad expects each set of sample data to be in adjacent columns In the example below we are testing if color affects our data measurement Select the region with the data lf Labels are included as in this example set labels to ON Set Data grouped by Column if the data is in columns as in this example Set the desired alpha level and tap Calculate CE One Way ANOVA One Way ANOVA Select the range of cels containing the Gata The factors should be grouped by columns or rows E m Alpha 05 First row column contains labes on Calculate ome Ba 1 CI c2 c3 c4 c5 C6 ENS P Ul A REO BLUE GREEN VELOWI 2 2 20 Ra 55 2228 54 5971
7. Tap the calculate button C Two Proportions Exampie icd ica Test For Difference in Proportions Uses normal apperosomabon X Number of successes observed N Total Sample sqe Mypotedzed Ovflerence O Nut Hypothesen on ERI o Agha 05 Calcul a e 47 Power Power of a Test calculates the probability of correctly rejecting the null hypothesis when the actual mean is different from the hypothesized mean by a specified amount Enter the hypothesized mean Enter the Actual also known as Stipulated mean Enter the Standard Deviation Enter the sample size Enter the Alpha value for the test The test uses a normal distribution by default To use a T Distribution this switch to ON Select the desired null hypothesis or alternate if Settings is set for alternate The Power will automatically be calculated and entered into the Power text field If a single power calculation is desired nothing more needs to be done Tapping Create Power curve will generate a curve displaying all possibly stipulated or actual means and the power of correctly detecting that mean Cub Powe one Ga Power of a Test CPUS AD Zul 0451 Mg 3 Ri Power 1 fi is te peobabity of correctly reyecting the m hypothesis amen fe actoal mean 6 atterent bom hypothesized Mypotheszed Mean 100 Actual Mon 110 Standard Deviation 29 Sampie Sze gt Aipha 045 Use T Dwutibuson OFE Nut Hypothess o E e Power 00099
8. This test can be performed using data points only For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button Set the confidence level If the range contains a label as in this example set the data label switch to ON Tap the Calculate button 21 Confidence Interval Variance Confidence interval for variance creates a range within which the true population variance has the corresponding probability of falling This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the sample Set the confidence level If the range contains a label as in this example set the data label switch to ON Tap the Calculate button e c 5 a2 Ar 21 or I od Ir 2 4i ad and und o p a Jala i mu 5 5 E 5 u 4 4 r E For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button 4 9 9 06 9 9 9 eo e oe 9 9 ce oe 9 oe o_o oe ooo 9 9 9 6 22 Confidence Interval Single Proportion Confidence interval for a single pro
9. another factor Common examples are before and after measurements where the person being tested acts as the other factor Use this test when a Paired T Test is desired but the data does not meet the criteria of normality Select the region containing the data for Sample 1 and tap the Sample1 button Select the region containing the data for Sample 2 and tap the Sample2 button Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button C Wilcoxon Paired Si Example iz i i Ei i a Wi ceren Metchad Selre E c7 ce co COO Signed Rank Test EE 1 l 39 ev ee 00 Nonparametric test lo Compare Medans of two s4pws ol data matched in pers Mypofedzed Oillerence 9 Nul Hypothesis tpl es mi j f E a Press to sol Current selectond Sample Paget H2L7TH 2C Senpe2 Paced R2CB R12C8 Fist tow oolumn corti bets ex Calcul a e 43 Normality Test A normality test determines if a sample of data deviates so much from a normal distribution that the data should not be considered as coming from a normal distribution This test uses the Anderson Darling method to calculate test statistics and p values for a set of data Select the region of data to be tested This region can span multiple rows and columns and will be treated as a single sample o
10. is greater han of equal to entered probabtty Probabiiny 00 Oceano keh Meo ml gi 1 Ave Peri 5 35 Means Test Z Test One Sample Mean Tests if asample of data came from a population with a hypothesized mean This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the data Set the hypothesized mean and the known Population Standard Deviation Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button CE e a L OOE AAA zz BE ss Fh Se Ee Se EP D L T Teel Dre Mean AI TSINO WOD PARED f u A L2 f i 1 9M fbit Phe arge of ced Confaemwg cw P Te ara A i i uo I Mystic Mea c B i ne f jez Pen Gs vem 3 E gt jam E i ul osse gt rim gs 4 NI Br ET W 4 R i41 E 7 ba AL i kore 4 7M 4 4 anm 4 fry ma rare miron QE For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Set the remaining fields as described for the Data Points section Tap the Calculate button CE C Z Tasi One Sampie Liampie 5 m m er WEN lo ER DEE TEE a os B
11. is the numerator degrees of freedom D O F 2 is the denominator Enter a probability to calculate an F value Enter an F value to calculate a probability EFF r Distribution m 3 Ga Le Ive I amp I cc I amp I amp F Distribution T Es be y i ni a bE a 54 Dip yD ra i JEJEJEJE PIT 4 4 ERE A gt 1 X i 4 s ke si 7 t o e UT ie au Emer degrees of freedom 1 and 2 Then enter a protabdly to calculate the F Value Erer a F Value to calculate he probabeny Probably 024 Vale 15645 DOF i 7 bor2 w 32 Exponential Distribution The exponential distribution will calculate a probability in the shaded area on the graph that corresponds to the X value Fill in the mean 1 A for the distribution Enter a probability to calculate a X value Enter a X value to calculate a probability Ce Exponemial Distrib Exponential Distribution Enner the population mean Then enter a probabtty to casa X Erer X to calcule the protatity Probably 0073 x Mean 1 Ak 10 cone Boag n con _ ga a 33 Binomial Distribution The binomial distribution will calculate a probability in the shaded area on the graph that corresponds to the number of successes Fill in the mean probability of an individual success P success Fill in the sample size Enter a number of successes
12. of data output in a matrix format The data needs to be grouped in adjacent rows or columns Select the range containing all of the variables of interest Select how the data is grouped by Column in this example If a T Test is desired on the correlation statistics set the Alpha value and set Include T Test to ON This will create T critical values and T Test statistics assuming a two tailed test for correlation If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Correlation R2 1110000 1000 Lxanmn Emp EN AC VD ERE d epi Ed wm nme P re cs C8 Tee Sq teet Age Bedrooms raras Garage t R3 133500 1400 i T gt a 1248 1106 2112 1078 at eon de b ho 952 B 3 INS IM Porr co L3 ome mre re gr 1100 1416 1150 DAC AIN 1 3 amp IL ic ah m A i IP alata 1220 tid td lea 160 oe Gol fg G2 i652 its 4 t 3 59 TRO TROT PO N AD amp 77 250 gt de Er 15S o o y 1A 1A AD A 25 Covariance Covariance calculates the covariance matrix for two or more sets of data The data needs to be grouped in adjacent rows or columns Select the range containing all of the variables of interest Select how the data is grouped b
13. test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button Enter the hypothesized mean difference Enter the standard deviation for population 1 and 2 Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Eo sae RA ZA ay Bao Soe moe P Test Two Meares umet TWC WARED teure we Mo Ac Teaser Fo J 7 Dersi A a anh o EE een ar ENE n al AL Pees u wd erred adam aor rt 9 9 49 349 4 4 4 For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button Carrier 100 m Z Test Two Sample Example Undo m Z Test Two Means Hypothesized Difference 0 Pop 1 Std Deviation 4 Pop 2 Std Deviation 5 Null Hypothesis BEE mq 1 1 1 Alpha 05 f pa Data Points Sample Statistics Sample 1 Sample 2 X bar 37 2 43 1 N 10 10 Calculate Analyses 41 Mann Whitney U Test A Mann Whitney U Test determi
14. to calculate the corresponding probability Enter a probability to calculate the corresponding number of successes Entering a probability will calculate the smallest number of successes who s cumulative probability is greater than or equal to the entered probability Ce Binomial Distribution Example icd idi Ca Binomial Distribution at t ut Ermer parameters nid successes Y cmalate cumulative probabey Pix s Successes Emering probabity returns smallest success who s comuawbve probabuy greater thao or equal w ectered probabtey Probabiny 005 Suctessaa 14 Sampie Sizo 100 Pi success 34 Poisson Distribution The Poisson distribution will calculate a probability in the shaded area on the graph that corresponds to the number of occurrences Fill in the average number of occurrences per unit measured Fill in the number of units measured Enter a number of occurrences to calculate the corresponding probability Enter a probability to calculate the corresponding number of occurrences Entering a probability will calculate the smallest number of occurrences who s cumulative probability is greater than or equal to the entered probability Ce Poisson Distribution Example idi Da Pe RE i EG A ST 00 09409 1 q Polsson Distribution Ereren Emer parameters ad Gccurrences amp calculate probabtty Pix s Occurrences Emering probabity returns smallest occurrences whos cumulative probab ty
15. y ne us 4 4 ewes m wet ner 4 i i 4 li For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button Carrier gt 6 49 PM 100 m Pop Variance Kno Example undo E p Confidence Interval For Difference In Two Means Assumes polulation variances are known Pop 1 Std Deviation 5 Pop 2 Std Deviation 5 Confidence Level 0 95 f panses Data Points Sample Statistics Sample 1 Sample 2 X bar 33 1613 42 4069 N 10 10 Calculate Analyses 19 Confidence Interval Assume Pop Variance Equal two means Assume Pop Variance Not Equal two means This analysis creates a range within which the difference between the true population means has the corresponding probability of falling Both Assuming Pop Variance Equal and Assuming Pop Variance Not Equal are setup the same The only difference is in the algorithm used to calculate the test This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button Set the confidence level If the range contains a label as in this example set the data label s
16. 0 0623 62 766 R7 40 063 H8 129 4168 58 6977 74 286 H9 703393 48 8742 57 938 RH10 42 7929 59 6096 62 975 RI 72 2714 R12 69 965 H13 55 7092 32 1486 62 262 Bj Analyses m ANOVA Chi Square Tests Confidence Intervals Correlation Covariance Descriptive Distributions Means Test iPhone iPod Windows Analysis Menu Contains menu options for all analysis file operations settings and help Data Table Contains data points to be used for analysis Data opened from Excel csv or tab delimited files will be placed in this window This is also the area to manually enter data Page Control Bar Shows how many pages are in the current file as dots and lets the user change pages by tapping on either side of the bar iPhone iPod Stats Pad on iPhone or iPod runs exactly like on the iPad with the exception that either the Data Table or the Analysis Menu are displayed at any one time Users on iPhone or iPod will use the Data Analysis button at the top right corner of the display to flip between these two window sections For example users may have to switch to the data table to make a selection and then switch back to the analysis menu to set that selection for the analysis Moving Within a Page There are often more cells in the data table area than can be shown on the screen To move around on a page to view cells not currently shown touch table area with one or more fingers and drag to the are
17. 901_ s22006 54 7966 56 1951 78 0513 20 DPUNCCUNUDIN he EAN R6 1D 160 063 40 0623 62 7666 560577 R8 IF 1294168 58 6977 74286 528197 ENS 7 p o pe ERE UN EN g Alpha 05 First row column contams labeis on Calculate 12 Two Factor ANOVA A Two Factor ANOVA also known as a Two Way ANOVA with Replication tests if multiple samples came from populations with the identical means where two factors are being varied There are multiple data points per combination of settings for the two factors see the example below This test is typically performed when two factors are set to multiple levels to determine if either factor affects the mean of the distribution For this type of test interaction between the two factors can also be tested For a Two Factor ANOVA without Replication see Random Block ANOVA Stats Pad expects each set of sample data to be in adjacent rows and columns It is also expected that the row factor levels are grouped together and have the same number of rows for each level In the example below we are testing if size and temperature affects our data measurement Select the region with the data Set the number or rows per sample number of replications In this example it is 3 Set the desired alpha level and tap Calculate lf Labels are included as in this example set labels to ON Cu Two Factor ANOVA Example L3 di LS Two Factor ANOVA
18. NOVA R2 RED BLUE c BE MEME NN Select ine range o cols containing M 552228 545971 51 0961 152 2496 EE EEE a ors shouid R4 28 108 547966 56 1951 780513 M EEE RS 02306 j5so4d 147 4219 158 19 _ a BEER AT EEE ne u RIS D X 1 O RDR A A aaa EE a MEERE E E NN First row ocolumn contains labes Cx Calculate 14 Chi Square Tests Contingency Analysis Contingency Analysis tests if two factors are independent This test is typically performed when there are two categorical variables being measured and the number of occurrences of each level of the variable can be counted Stats Pad assumes that a contingency table summary table of count frequency has already been constructed as inthe example below Stats Pad expects the data to be arranged in a contingency table In the example below we are testing if height and width are independent Select the region with the data Do NOT include labels for this test Set the desired alpha level and tap Calculate Gud Contingency Analy Example Une Rede Ld uc __ Ci c2 C3 j C4 C5 Ce C n2 QJWIDE JMD NARROW Fertig dicor sapo Ara RHs ISHORT Do 1155 t5 w independent of the column fact MO I 15 I Expected frequency data wil be RS jT O 15 9 pmo o J 0 jj normalized for analysis BE LS I LL ee md DO NOT include labels Contingency Analysis Analyze a c
19. STATS PAD USER MANUAL For Version 1 8 4 Manual Version 1 1 Table of Contents Basic Navigation Settings Entering Data Exporting Data Managing Files Running Tests Interpreting Output ANOVAs Chi Square Tests Confidence Intervals Correlation Covariance Descriptive Distributions Means Test Normality Test Percentiles Proportions Tests Power Random Numbers Regression Sample Size Variance Tests O CON N C 10 11 15 17 25 26 27 28 36 44 45 46 48 49 50 52 54 Basic Navigation Window Components u Gau ANOVA Am TU E R2 RED BLUE GREEN YELLOW Bicis sasori unen oie Cells not Se FMrjoB 05 154 7966 55 1951 780513 851302395 458948 474213 5819 recognized Correlation BS 150063 600623 627666 550577 as numeric Ariss osa2 600042 400533 757168 Covariance Ae sasa ener asse ws Will be gray fig 70 3393 488742 579382 635279 39T Fern Kigja27529 596096 629755 676558 Distributions BINS 5422 34996 722714 646045 J o Ki2is7o842 513781 699654 88457 o Mears Test mi1s57002 1321486 l622628 amp 25367 EE PUR MEM op Sc zA Percentiles Proportions Tests Data Table Analysis Menu Power Page Control Bar iPad Window pe cH Examp Menu RI ITWAY O R2 RED BLUE GREEN R3 51 0361 H4 28 105 54 7966 56 1951 RS 130 2395 45 8948 47 421 R6 160 063 4
20. Signed Rank test determines if a sample of data came from a population with a hypothesized median Use this test when a t test for a single mean is desired but the data does not meet the criteria of normality Select the region containing the data Set the hypothesized median Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Ca Wilcoxon Signed Exampie Da i Wilcoxon Signed Rank Test a e i Non parametnc test for ore sample medan lelg S alels C017 JI Mypothesized Median 15 P in h Un m 4 74 Fi Nut Hypothess BEER 5 at Din in 318702 1a1 3131 Aipha 05 Bizgpseor 1 2057 asra Do JOLENE 20 IHSENEBOS BEER HENENEU a EEG Y Fist rowloctuew Corta Wbets Cr Celculete 38 T Test Paired A Paired T Test determines if a two samples of data came from a population with a hypothesized difference in means where the data is paired The data for this test should be tied together in pairs by another factor Common examples are before and after measurements where the person being tested acts as the other factor Select the region containing the data for Sample 1 and tap the Sample1 button Select the region containing the data for Sample 2 and tap the Sample2 button Set the
21. a Fe Mura Lic rere omm feit age Of ced contenga me ur o_o 9 6 o gt e eo oe 9 9 9 9 9 9 9 9 9 P o 9 9 9 9 9 9 e ooo o 258 Fae ESAT 18 Confidence Interval Pop Variance Known two means This analysis creates a range within which the difference between the true population means has the corresponding probability of falling For this test Stats Pad assumes the population standard deviation is known This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button Set the known Population Standard Deviations 5 and 5 in this example Set the confidence level If the range contains a label as in this example set the data label switch to ON Tap the Calculate button GERM Pop Variance Kno Limpie nad bd ined iz ES d E me p E B 2 CM EN Be 8 Ceciderce hers m Pee DN wr o wo ry mex b Tee ee ais P9373 X7 ar 3 E LIA ur d Im Basar DOSE n nr amp Kae 120 129 c BALY 1 171 THOMAS 6 109 TOOY LIRIFIEI IEEE E i288 83rf68 0 9 9 9 9 9 9 9 9 9 9 9 9 4 t o n IM IB lt w
22. a you wish to view To avoid accidentally selecting a range of cells it is best to use two fingers to touch and drag the sheet Adding Deleting Rows and Columns Add or delete rows or columns from the end ofthe data table by dragging the row or column add delete buttons Analyses ANOVA m mS RED BLUE YELLOW TM Aa 155 2228 54 5971 51 0361 522495 T 7 ccce wee R4 28 105 54 7966 56 1951 780513 R5 130 2395 45 8948 arais eso lo o o Correlation R6 60 063 J R7 148 0842 Covariance Bio 2109586077 piae ees m 70 3393 529982 E 5279 H Descriptive R10142 7929 50 6006 Fu IE ern 6558 RI ls 164 604 Distributi ons A12ls7 9842 51 3781 89 9654 58 B4 Add Delete Columns Means Test Bis 221406 fee 2620 jet T 1 R4 O Ee Normality Test B 1 31 1 Percentiles Proportions Tests Add Delete Rows Power Random Numbers Regression Insert a row column in the middle of a data table by tapping on the header cell and choose Insert from the pop up menu Selecting Cell Ranges To select a range of cells touch a cell at the upper left or lower right corner of the desired range then drag to the opposite corner of the range If you are having trouble accidentally dragging the sheet when you wish to select a range pausing for a second after touching the first cell and then drag To select an entire row or column tap on the row or column header Drag
23. ailable Polynomial Order Only available if Polynomial is selected Force Y Intercept to 0 It is advised that this be left to OFF unless the user is very familiar with Regression Through the Origin and its potential drawbacks Confidence Level This will be used to create upper and lower limits for the coefficient estimates Output Residuals This will output a set of Residuals or differences between actual and predicted values This can be used for further tests such as residual analysis Select the range containing the Y or Dependent data and tap the Y Data button Select the range containing the X or Independent data and tap the X Data button For this analysis only one column or row of independent data can be selected The data can be grouped by row or column The test will automatically determine the correct setting based on the Y data selection If the first row or column contains labels as in this example set labels to ON Tap the Calculate button Functun Fit 4 a h3 A Iss In T J Jd IP T gt J _ _ 4 e a D J t d J J i zb Tap AAD AD 4 4 9 gt Un i Ge Wel wi J hj 9 b e EE En An ee Ge GENE GAN Gn HERE HEHE Ga GENE Gn MENU en Gee GE Gann Ge Gin Gi HERREN GEA HERREN 51 Sample Size Sample Size for No
24. ample Size will automatically be calculated and entered into the sample size field This test rounds UP on sample size to ensure you have a margin of error equal or smaller than the desired margin of error C Proportion Sample Size Proportions Enter desred margin of error populabon proporion ad contiience wn to catculete equred sample size NOTE Sample szo wi round ue Magn of Error o Popudanson Proportion 04 Contdenos Level 0975 Sample Size 53 Variance Tests Chi Square Single Variance Tests if asample of data came from a population with a hypothesized variance This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the data Set the hypothesized variance OR standard deviation Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button 2 8 u eee on han und lad nd Mit er Pe Se Sa ee Ch pam Teel f nery LE tav einge te namen A TSIN TWO PARED H2 DAMME EAMAN ze N u F7 PERSON BEP re rg ren mr jns 1007008 JOMN io mT je teps i50 m ss Im 366 e Ld 4 ev bao H 23 9367 je 1218 H wo 41 431 BEP M9 1405501 Ped tym dne e
25. f data Enter the alpha level for the test Tap the Calculate button CSF Normaiity Tes Exampie e n p UT GN Poe pP ge Up d OM Ine Normality Test EEE I 1 1 R2 RED BLUE GREEN vElLOW Tost if curenty selected dala toliows RSS 2228 545971 51 0361 522496 a normal distribution using the Bazjs 105 154 7966 56 1951 780513 OO HG Anderson Daring test Non nomenc data wi be inored 827666 1560577 Aipha 05 40 0633 75 7168 74206 528197 579382 635279 fans 64 6045 ENG GENE 58 8457 iSite lb ASIE ale PM m p _ Calcul a e 44 Percentiles Percentiles will calculate percentile values for all deciles increments of 10 and quartiles increments of 25 as well as one additional specified percentile Percentiles will be calculated on all data treated as one sample as well as each column or row depending on the Data grouped by setting Select the range containing the data Set the Data Grouped by columns for this example Set the additional percentile if desired If the first row or column contains labels as in this example set labels to ON Tap the Calculate butto C Percentiles Percentiles Select the range of celis containing n te data Al quartes and decies wil be calcuated along wih Pe speched percento Data grouped By id Mur om Percentie First row oolumn contis labels Calculate
26. ging to another row column header will select multiple rows or columns To select the entire data table tap the button in the upper left corner of the data table Tap again to deselect any selected ranges To modify a range touch the upper left or lower right cell in the range and drag to the new start or end cell To hide or display the copy paste menu for a selection single tap inside the selection Changing Adding and Deleting Pages To change the currently visible page tap the page control bar on the right or left sides To add a page tap the Page button at the top right corner of the data table window and select Add Page To delete a page tap the Page button and select Delete Current Page Resizing Columns To manually size a column touch and hold the column header near the right edge Drag right or left to increase or decrease the width Text will automatically adjust to the new width To automatically size a column select the entire column or columns From the copy paste menu select AutoSize Settings Some analysis behavior can be customized by tapping the Settings option on the analysis menu Currently there are two settings available Choice for Hypothesis Tests Determines if the null or the alternate hypothesis choices are presented when running hypothesis tests The default is the null hypothesis Some text books teach that the null hypothesis will always contain not or gt For compatibility with
27. hypothesized mean difference Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Ca T Test Paired Exampie iz Da T i d Paired T Test Ai PARED Mosis 2 PERSON BEFORE AFTER Compares means of two sarngtes of R3 JOHN 140 Ly dala matched e pars Besa q2 IE 6l eere MENSIS un 1 OR m jweENDY fioo Pio Jo r Rd coco o RIOIKATE 105 Bd An ae m NN Oe R12 p H7 jj bl O D a a mm SS a a as E wm 8 11 1 1 LR A Press to sol Corte selectors Sample Paget H2C7THWRH 2C7 Senpi Paced R2CBRI2CB Fist tow oolumn contains abes CELE Calcul a e 39 T Test Pooled Variance or T Test Separate Variance Test if two samples come from populations which have the hypothesized mean difference Both Pooled and Separate Variance are setup the same Use Pooled Variance if it is believed the populations have the same variance Use Separate if it is believed the populations have different variances Ensure D O F Rounding is set to the desired preference in Settings This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select
28. ic file name will reveal a Delete button Tapping the delete button will permanently delete that file Renaming Files Tap the Edit button at the top right corner of the file menu window to place all files in edit mode Double tap directly on the name of a file to place the file name in edit mode When done editing tap the dismiss keyboard or the return key GUEd EA New File Example Running Tests General Single Sample Navigate to the desired test and choose the appropriate settings Navigate to the data page and select the range of data for the test Tapping the Calculate button will run the test on the currently selected range of cells Output from most tests will be placed in a data table on a new page Two Samples Navigate to the desired test and choose the appropriate settings Navigate to the data page and select the range of cells containing the appropriate data for sample 1 and tap the Sample1 button The sample 1 label will now reflect the chosen range Navigate to the data page and select the range of cells containing the appropriate data for sample 2 and tap the Sample2 button The sample 2 label will now reflect the chosen range Tapping the Calculate button will run the test on the chosen cell ranges To revisit either sample ranges tap on the sample label in the analysis section The data table window will automatically change to the appropriate page and select that range of cells Interpreting Output Stats
29. lace the tables found at the back of most textbooks Output will be displayed on the analysis screen not in a data table The following pages describe each distribution 28 Normal Distribution The normal distribution will calculate a probability in the shaded area on the graph that corresponds to the z or x value To use the standard distribution u 0 o 1 select the Standard tab on the selector Enter a Z value to calculate a Probability Enter a Probability to calculate a Z value Qum Normal Dist inanon Cee LL 4 EB M 5 4 4 Y i i 4 4 i no T Eros a poceni k catotete me Ex ZT was Cree a 7 un 1o conar te prote ga LOS Io vw abt s To use a distribution with a mean and standard deviation other than the standard select the Non Standard tab on the selector Fill in the mean standard deviation and sampe size Enter an X value to calculate a Probability Enter a Probability to calculate an X value nun Normal Dist iaon m amp e e ooo e e e o o 6 6 _ 6 _ 6 _ _ o e9 e9 e9 e9 e e e e e free ppasier res wgra ovi Fi BE 1 119 5 8 ao lsd 9 he ca cnm Pe orca Steed Pop Wee WW P Say 3 Sere Due 29 T Distribution The t distribution will calculate a probability in the shaded area on the graph that corresponds to
30. med using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the data Set the known Population Standard Deviation 2 5 in this example Set the confidence level If the range contains a label as in this example set the data label switch to ON Tap the Calculate button For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button 17 Confidence Interval Unknown Sigma single mean Confidence interval for a single mean with unknown sigma creates a range within which the true population mean has the corresponding probability of falling This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the data Set the confidence level If the range contains a label as in this example set the data label switch to ON Tap the Calculate button SL Lise ruens op eee inge of ced Contiene fe we i ed 2 d d d ud E n Nl pe 1 H ne s Md 9 4 9 34 For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Tap the Calculate button CE S Qd Ur cc cw boma Campie m bead 8 m E es Ss a es Bs a PE 4 ox or
31. nes if two samples come from populations which have the same median Use this test when a t test for two mean is desired but the data does not meet the criteria of normality Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Claudi Mann Whitney U Test Mann Whitney U Test Non parwmetnc teat lo Compare medians of two undependent surchs These test wel output lower tal crica wad paues Nut Mypothess Press to sol Curre t selectord Sample Paget H2CL3H 2CJ Senpi Paget R2C4 FH 12C4 Fist tow oolumn cort ari Wwbets Cr Calcul a e z AMPLE TISAMPL Aoo o ERSON BEFORE AFTER 259367 386 1218 t 120 11667 405501 WENDY ioo 9 321785 574723 RIC eo 172 sna ors ERI 19s 31 8702 M13131 MIKE 1145 135 443734 MOUENES OO 7 E Example iz ga 53 i3 35 un j M SESEEEEEEEREEEREEES i 42 Wilcoxon Paired Signed Rank A Wilcoxon Paired Signed Rank tests determines if a two samples of data came from populations with the same median where the data is paired The data for this test should be tied together in pairs by
32. nm 1 2771 wires 1574723 E Han tz 1 Bae mob pes 431995 LETS D tbe jn a j 3 71 ld 4 kor mi 42 M 534 JM 13 ina 4 4 4 4 4 DT C AD 4 4 4 4 4 ine LARA 4 4 4 4 4 4 4 4 4 4 For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Set the remaining fields as described for the Data Points section Tap the Calculate button Chi Square Single Example Chi Square Test for single sample variance Tests variance of a single sample Hypothesized Variance 9 Hypothesized Std Dev 3 000000 Null Hypothesis Var s Hyp BEE Var Hyp Alpha 05 D Data Points Sample Statistics Sample 1 Std Dev 5 717 N 10 Calculate Analyses 54 F Test Two Variances F Test for two variances test if two samples come from populations which have the same variance This version of F Test will arrange the samples so that the upper rejection region is the only region of concern i e rejecting the null hypothesis will only occur in the upper tail by placing the appropriate variance in the numerator This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the first sample and press Sample1 button Select the region containing the second sample and press Sample2 button
33. o create upper and lower limits for the coefficient estimates Output Residuals This will output a set of Residuals or differences between actual and predicted values This can be used for further tests such as residual analysis Select the range containing the Y or Dependent data and tap the Y Data button Select the range containing the X or Independent data and tap the X Data button The data can be grouped by row or column The test will automatically determine the correct setting based on the Y data selection For simple regression only select one X variable For multiple regression select more than one X variable If the first row or column contains labels as in this example set labels to ON Tap the Calculate button Seer TIT oe Simple amp Mulole Regression b 74 z1 no 3 gt T t en 9 a Tro a Tz 3 JIMS 4 e 4 ADI AD JAD ae i ae an P 4 9 WIENER we m I A A Un Ge wl w 1 J a I I I gt gt gt 5 1 Do S MIESIE PIS IPFO r core r 7 50 Function Fit Function Fit regression will calculate a least squares fit model that predicts Y values based on input X values which are used in a selected nonlinear model This test will also provide all of the statistics necessary to determine ifthe model is valid and significant The following settings are av
34. ontingency table to 15 Goodness of Fit Goodness of Fit tests if the observed frequency matches an expected frequency This test is typically performed to see if a sample set of data matches an expected or theoretical distribution Stats Pad assumes that an expected frequency and observed or actual frequency table has been constructed The Expected frequency can be either absolute numbers or relative frequency percentages In the example below we are testing the observed number of M amp Ms matches our expected frequency of M amp Ms Do NOT include labels for this test Select the region with the Observed or Actual data and tap the Observed button Select the region with the Expected frequency and tap the Expected button Set the desired alpha level and tap Calculate GEESE Goodness OF r compe Gau C e ca ca ca es ces I c RER Goodness Of Fit in 1 Perform Goodness Of Fit test of a sample against an expected frequency Expected frequency data wil be normalized for analysis DO NOT indude labels Jesse Alpha 05 Observed Pages RGC3 R8C3 Expected PageS RGC2 R8C2 Calculate 16 Confidence Intervals Confidence Interval Known Sigma single mean Confidence interval for a single mean with known sigma creates a range within which the true population mean has the corresponding probability of falling For this test Stats Pad assumes the population standard deviation is known This test can be perfor
35. portion creates a range within which the true population proportion has the corresponding probability of falling This test can only be performed using calculated statistics Enter X with the number of successes observed success is defined as finding the value of interest Enter N with the total observations Successes and failures Tap the Calculate button Qoi Single Proportion Example icd idi Ca IT 01 SC aed a ae ey es A Confidence interval For Proportion f Ru TA OR MS u Uses normal approsmaton X Number of successes observed N Tote Sample sve Confidence Level 295 Calcul a e 23 Confidence Interval Two Proportions Confidence interval for the difference in two proportions creates a range within which the difference in the true population proportions has the corresponding probability of falling This test can only be performed using calculated statistics For each sample Enter X with the number of successes observed success is defined as finding the value of interest Enter N with the total observations Successes and failures Tap the Calculate button Coal Two Proportions Confidence Interval For Difference in Proportions Uses normal approsomabon X Number of successes observed N Total Sample sine Conhdence Level 295 Sarrygte 1 Sample 2 x 4 43 Calcul a e 24 Correlation Correlation calculates the correlation between two or more sets
36. py Paste Moving data to another app can be accomplished using copy paste Select the range of cells t0 be copied and use the copy function Navigate to the app where you wish to paste this information and use that app s paste function Only apps that can paste tab or comma delimited information will recognize the data File sharing The currently visible data table page can be exported as a file to another app by tapping the share button in the upper right corner of the data table window Choose to share as either acomma or tab delimited file and the apps which recognize that type of fill will be displayed for sharing Select the appropriate app and the fill will be sent to that app Managing Files All file management functions are accessed from the Files and Settings tab by tapping the files button Opening Files Tap on the name of the file The currently open file will be saved prior to opening another file Creating New Files Tap the button at the upper right corner of the file menu window Enter the name of the new file when prompted Tap the Create button to finish creating a new file Deleting Files Files can be deleted in one of two ways Swiping from right to left across the name of a file will reveal a delete button Tapping the delete button will permanently delete that file Tap the Edit button at the top right corner of the file menu window to place all files in edit mode Tapping the red button for a specif
37. rmal Distribution This analysis will determine what sample size is required in order to achieve a given margin of error for estimating the population mean It is assumed that the population standard deviation is known through testing or other means Enter the desired Margin of Error Enter the Population Sigma Enter the confidence level that would be used for calculating the margin of error The Sample Size will automatically be calculated and entered into the sample size field This test rounds UP on sample size to ensure you have a margin of error equal or smaller than the desired margin of error C Normal Distribution Sample Size Normal Exampie bandi buch na EL b Enter desred margin of error populaton ugma and contience love to calculate reqgured sarpi sve NOTE Sampie sizo wi round up Margn of Error X Populwbton Soma S Contdence Level 095 Sample Size 52 Sample Size for Proportion This analysis will determine what sample size is required in order to achieve a given margin of error for estimating the population proportion Itis assumed that and estimate of the population proportion is known through testing or other means If unsure of an estimate to use enter 0 5 as this will give the maximum sample size necessary Enter the desired Margin of Error Enter the Population Proportion estimate or 0 5 Enter the confidence level that would be used for calculating the margin of error The S
38. s HAE E Du T eel Dre Mara 6 i i n ete Pe tage Of Com coorta g mE rom he 4 4 i 4 4 i EN fon n NOT e i i i i i N m Feen yet Canvas 7 E f f E 4 4 4 i 4 a A Heuer e 4 4 4 4 4 4 v NL co o 36 T Test One Sample Mean T Test for a single mean tests if a sample of data came from a population with a hypothesized mean This test can be performed using data points or pre calculated statistics For data points select the Data Points section of the selector Select the region containing the data Set the hypothesized mean Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button T ei Dre Ma fete age of ced COE fe ui For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields Set the remaining fields as described for the Data Points section Tap the Calculate button Quim Tes O Gatipte Liampie DD i3 G I C T ei Dre Mea Zara ange of cef Diarog re Ow a 9 9 9 9 9 9 9 9 gt 9 e6 e 9 9 4 37 Wilcoxon Signed Rank A Wilcoxon
39. the region containing the second sample and press Sample2 button Enter the hypothesized mean difference Select the desired null hypothesis or alternate if Settings is set for alternate Set the alpha level for the test If the range contains a label as in this example set the data label switch to ON Tap the Calculate button hen e a 14 1 a ILI 9 I fa A77 Ar Lo SOA l LEO mr te ace u rg i eam gt M poc ALLY L ES MOMAS E ony Tew Lm om E 721 En MOY Mai heyy nan E om i E 2773 s oom e t m no me J ew i i zn I NI P 1 IN MED 13 i t 14 alm 15 gt gt e t eepe eoe un u u 9 9 9 9 9 For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields and tap the Calculate button Carrier gt 9 43 PM T Test Pooled Varia T Test Pooled Variance Assumes population variances are equal Hypothesized Difference 0 Null Hypothesis 1 2 s Diff EE a u1 u2 2 Diff Alpha 05 Sample 1 Sample 2 X bar 35 2 40 3 Std Dev 4 326 5 012 N 10 10 Calculate Analyses 40 Z Test Two Sample Means Tests if two samples come from populations which have the hypothesized mean difference For this test it is assumed that the population variances are known This
40. the t or x value To use standard distribution u 0 o 1 select the Standard tab on the selector Enter a T value to calculate a Probability Enter a Probability to calculate a T value G 1 siren amp e e e e e ooo oe ce 9 9 9 e eo gt EEE EEE To use a distribution with amean and standard deviation other than the standard select the Non Standard tab on the selector Fill in the mean standard deviation and sampe size Enter an X value to calculate a Probability Enter a Probability to calculate an X value mn T Distrauton nnp p e oe o o o o o o o 30 Chi Square Distribution The chi square distribution will calculate a probability in the shaded area on the graph that corresponds to the chi square value Fill in the degrees of freedom D O F for the distribution Enter a probability to calculate a chi square value Enter a chi square value to calculate a probability Cd Chi Square Distrib CTPUSDUTTEEUTTUEEUTTCEEUTTE Chi Square Distribution EI 2 C3 C4 Erner a probability to calculate Chi Square Erner Cte Square t calculate the probaberty Probabiiity 0 05 Chi Square 163070 DOPF w 31 F Distribution The F distribution will calculate a probability in the shaded area on the graph that corresponds to the F value Fill in the degrees of freedom D O F for the distribution D O F 1
41. these text books change this setting to alternate hypothesis Fractional D O F Rounding For tests involving two sample T tests assuming unequal separate variance the calculation for degrees of freedom can result in a fractional number Choose the desired option Nearest uses standard rounding rules to round to the nearest integer Down will round to the next lower integer Up will round to the next highest integer Entering Data Manual Entry Data can be manually entered by double tapping a cell Pressing the return key when finished will enter the data and move to the cell directly below When finished tap the dismiss keyboard button or single tap another cell Copy Paste If data already exists in another app on your device such as Numbers select the data in that app and use the copy function to place that data on the pasteboard Switch to Stats Pad and select the upper left corner cell where you want to paste the data and choose the paste function The data table will expand if necessary to accommodate the new data Excel CSV or Tab Delimited Data from an Excel xls comma delimited csv or tab delimited txt file can be opened directly into Stats Pad While viewing the file in mail or other app such as Dropbox use that app s Open In function to open the file in Stats Pad Only data information will be imported to Stats Pad Graphs pictures or other types of information will be ignored Exporting Data Co
42. witch to ON Tap the Calculate button inti zu zo um 4 I C ae EM So EN WIRED c AMPLE SAMALE F PERSON BEFORE AFTER BT PORON r dec an m D oos JOMN 0 100 rn nn BALY 17 Im n THOMAS 7 6 jm MON x m n JRE je 153 wc wen io vr RR ieo IK mo KATE is jos mt te MKE its 1125 a aa mz 222 7 3734 JOLENE Im T UAE a AT N us i 4 4 eres m uni over Samum tur gt i b 4 4 For pre calculated statistics select the Sample Statistics section of the selector Fill in the sample statistics fields and tap the Calculate button Carrier gt 7 18 PM 100 mm Assume Pop Vari Example cS 1 Ct c2 c3 c4 C5 Ce C HEN MEMINI Confidence Interval For Difference In Two Means Assumes population variances are equal Confidence Level 0 95 Data Points Sample 1 Sample 2 X bar 33 16 42 41 Std Dev 4 508 5 746 N 10 10 Calculate Analyses 20 Confidence Interval Paired Sample Data Confidence interval for difference in two means where the data is paired creates a range within which the difference between the true population means has the corresponding probability of falling The data for this test should be tied together in pairs by another factor Common examples are before and after measurements where the person being tested acts as the other factor
43. y Column in this example Set ifthe data represents a sample or the entire population Sample in this example If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Covariance Data grouped by Lec ster pr mn LH ta iGo ae te AD A A i PA sha to te tal UU 3 9 IPO TIPO ADP A it iG 57250 PIRIN m INN a iGo 15215 152500 RIR 7 26 Descriptive Descriptive Statistics calculates many of the common statistics on a set of data This function will calculate a grand total set of statistics as well as on each column or row if grouped by rows Select the range containing all of the data of interest Select how the data is grouped by Column in this example If the range contains a label as in this example set the data label switch to ON Tap the Calculate button Descriptive Statistics 52 2496 Calculate Gesorptive stabs cs on Pe 780513 currerity selected data 62 9755 67 6558 Data goued by ii m Bk S122 sa oe 22714 amp 645 O Jo O seo fer 3761 essem je Fist rowlootue cort ra Wwbets ex Cak ulate 27 Distributions The distributions section provides a quick means of finding the critical value for a corresponding probability or a probability for a corresponding critical value These analyses rep
Download Pdf Manuals
Related Search