SPSS Practical Manual on Normality Checking of Data: A Guide to Testing Data Distribution, Summaries of History

Download SPSS Practical Manual on Normality Checking of Data: A Guide to Testing Data Distribution and more Summaries History in PDF only on Docsity! SPSS Practical Manual on Le) OT) AUG 1.0 4 H=0, O7=0.2,—| =0, 0?7=1.0,—— a \ “=0, 07=5.0,—— a M=-2, 07=05,——=| 4 —~ 06 | J % S 04 / 0.2 1 N Lg a —_ 4 S _— 0.0 a 11 1 J ie al ro aa J ai 1 5 -4 -3 -2 1 0 1 2 3 4 F iene? Ea Fil hl raraana D. S. Dhakre, D. Bhattacharya and Bhola Nath Institute of Agriculture, Visva-Bharati, Sriniketan West Bengal -731 236, India SPSS Practical Manual on Normality Checking of Data D. S. Dhakre, D. Bhattacharya and Bhola Nath Institute of Agriculture, Visva-Bharati, Sriniketan West Bengal -731 236, India Example: The following table gives the yields in pound per plot, of seven varieties of a crop after being applied to each of 4 plots, tested in a Completely Randomized Design. Test data is normal or not treatment RL R2 R3 tl 17.8 24.4 18.1 t2 20.2 22.7 23 13 15.7 16.4 18.8 t4 19.7 15.3 19.8 tS 19.4 20.3 23.7 t6 17.7 18.6 21.1 t7 21.7 21.6 17.4 There are three methods to check Normality of Data 1) the Z value of Skewness and Kurtosis should be -1.96 to +1.96 2) Shapiro Wilk’s test significant value should be above 0.05 3) with the help of Histogram, Q-Q plot, Box plot Arrangements of data for SPSS Rep Trt yld 17.8 20.2 15.7 19.7 19.4 17.7 21.7 24.4 Nielelefelelale RINT DOB] QO) rp] SESS Practical Manual on Normality Checking of Data Step 4: Analyze > Descriptive Statistics > Explore > Tesh nn TP FE oe a cea a — replication | weulinen || Senerett near te=ne are Ecrosttn. TURE eens Ham Wa er rts oan. Reatcion apres cy Nonparerst Tests crscening Buea ‘unpte esponse FE mutton tay ca , Fao cne pain croTomper Meany av SE iks)2)o)@|¢ |= es a DG! els) Step 4: Analyze > Descriptive Statistics > Explore — Select dependent variable eat of Neva (Dall - BM SPSS Sats Date ES See ote Usa transom Anajcs raphe Gusts UnesAdoars Wrenn 20 cs BLS a B= BOM o® *s) 1 Wate aoa velu [resteaon [ueanee [ved [er we [we ee ee ee ee 370 sa 4770 270 2a 2270 1640 Wwf x = (Go) (ee [esse )[Ceneat) Hen] ee SeeSorsnae =e | eee SESS Practical Manual on Normality Checking of Data Step 5: Plots..+Normality plots with tests > OK (8 "Test of Re-vsltysr [Datasecll IDM 5°25 Statistics Data Elite (6lel SAG Ros Ahae he RAG oe : al cons T ee a | = = feat Ss we] « 2 ae fon oa ene Lemont Tot : 4 — 45: la - : ‘eid 561s 6 = Ons Aone S Ll aaa remem Output: Descriptives Statistic |Std. Error Mean 19.9833] .60143 Lower Bound | 18.7144 y fi | | for M 95% Confidence Interval for Mean Upper Bound] 21.2522 5% Trimmed Mean 19.9981 Median 19.7500 Variance 6.511 yld Std. Deviation 2.55164 Minimum 15.30 Maximum 24.40 Range 9.10 Interquartile Range 3.95 Z value of skewness = 0.026/0.536= 0.049 Z value of Kurtosis = -0.722/1.038= -0.696 5 SESS Practical Manual on Normality Checking of Data Tests of Normality Tests of Normality Kolmogorov-Smirnov? Shapiro-Wilk Statistic df Sig. _| Statistic df Sig. yld 084 18] 200° .983 18 *. This is a lower bound of the true significance. a. Lilliefors Significance Correction Hypothesis under Shapiro-Wiilk’s test statistic is defined as: Ho: The data are normally distributed Hi: The data are not normally distributed According to Shapiro-Wilk’s test result is non-significant it means, we cannot reject Ho it means data are normally distributed Histogram Mean = 19.98 Std. Dev. = 2.552 N=18 Frequency yld Histogram should have the approximate shape of normal curve.