Several methods for determining normality using SPSS

Question: data on the number of male red blood cells (blood) in a certain place (seeProgramData between "begin data" and "end data") is the data normally distributed? (Introduced by Ma bingrong's master book, 2001. p.150)

The syntax window encoding is as follows:

* ma bin-Rong: SPSS for medical application, 2edit. 2001, p.150 :.
data list free/blood.
Begin data.
568 460 500 580 560 434 561 570 519 645 563 552
540 541 461 501 581 620 573 518 562 597
551 574 480 481 542 462 502 584 517 637 580 547 521 442
564 575 482 543 463 503 585 572 541 525
495 523 634 532 565 483 544 464 504 559
587 494 522 595 577 484 545 558 505 493 586 622 524
456 576 527 490 579 557 546 466 506 572 533
450 566 528 491 567 556 465 485 547 588 507 589 535
596 492 569 555 578 513 530 486 548 534 508 588
628 526 554 531 512 570 514 521 487 459
end data.
NPAR tests/K-S (normal) = blood/Statistics = descriptives.
descriptives variables = blood/Statistics = All.
frequencies variables = blood
/Statistics = all
/histogram = normal.
pplots/variables = blood/type = Q-Q.

Note: "* '''." In syntax, it indicates the meaning of the comment.

The frequencies statement is used to describe the frequency; the NPAR tests/K-S statement is a K-s test to check whether the data conforms to a specific distribution. Oraml indicates the normal distribution of the test. In addition, the uniform distribution is uniform, the Poisson distribution is Poisson, and the exponential distribution is exponential. The pplot statement generates a PP diagram to check whether the normal distribution is normal.

 CodeThe result is as follows:

 

 

Result Analysis:

In the result of Kolmogorov-Smirnov test, the Z value is equal to 0.532, and the P value is equal to 0.940> 0.5. Therefore, the data is approximately normal.

In the descriptive statics results, the skewness coefficient skewness =-0 .. 33; the kurtosis coefficient kurtosis =-0.517; both coefficients are less than 1, which can be considered to be approximately normal distribution.

In QQ plot, each point is approximately a straight line, indicating that the data is approximately normal distribution.

Result thinking:

There are many methods to test the normality of data. K-s test, descriptive, QQ plot and other methods can all test the normality of data.