full2010.pdf - page 641

603
Abstract
The purposes of this study of One – way ANOVA were to investigate the power of the test of One – way
ANOVA after transforming with large sample size data by using Real Data and used 5 sample sizes of 30, 60, 90,
120 and 150 students to saw if there were any difference among the tests and which method yields to the most
suitable result and at which sample sizes. The samples consisted of 2}400 Mathayomsuksa in Thailand by using
the Multi – Stage Random Sampling. The instrument for the study included the questionnaire of democratic child-
rearing that score reliability were 0.719.
The finding were summarized as follows:
1. when the sample size was increased, power of the test was increased.
2. when the data transformation was converted from square – root transformation, logarithmic
transformation, reciprocal transformation to raw data, power of the test was not different.
3. when the significant-level was increased from .05 to .10, power of the test was increased.
Keywords: Power of the Test, One Way Anova, Large Sample Size
‡Î µœÎµ
‡ªµ¤œn µ‹³Áž}œÄœ„µ¦Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ °œž¦³Á£šš¸É 1 Á¦·É¤˜oœ¤µ‹µ„„µ¦ª·‹´¥šµŠ—oµœ„µ¦Â¡š¥r Ž¹ÉŠ
®¨¸„Á¨¸É¥Š„µ¦Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 1 Ánœ „µ¦š—°¥µ¦´„¬µÃ¦‡œ· —Ä®¤n „µ¦„ε®œ—¦³—´œ´¥Îµ‡´
šµŠ™·˜·˜o°Š„ε®œ—Ä®o¤¸œµ—Á¨È„š¸É »— Ž¹ÉŠ‡ªµ¤œn µ‹³Áž}œÄœ„µ¦Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 2 ‹³™¼„¡·‹µ¦–µ
¦°Š¨Š¤µ Ĝ„µ¦ª·‹´¥šµŠ—oµœ¡§˜· „¦¦¤«µ˜¦r ¨³´Š‡¤«µ˜¦rÁž}œš¸É´—Á‹œÂ¨o ªªnµ œ´„ª· ‹´¥˜o°Š„µ¦®¨¸„Á¨¸É¥Š„µ¦
Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 1 Ánœ„´œ ×¥šÎµ„µ¦„ε®œ—¦³—´ œ´¥Îµ‡´šµŠ™·˜·š¸É¦³—´˜Éε °¥nµŠÅ¦„Șµ¤¤¸Ÿ¼o
Øo Â¥oŠªnµ „µ¦Á o¤Šª—˜n°„µ¦„ε®œ—¦³—´ œ´¥Î µ‡´šµŠ™·˜· °µ‹nŠŸ¨Ä®o Á„·— o°Ÿ·—¡¨µ—‹µ„Ÿ¨„µ¦š—¨°Šœ´ÊœÅ—o
®¤µ¥‡ªµ¤ªnµ ˜´ªÂž¦°· ¦³°µ‹¤¸°·š›·¡¨‹¦·Š˜n°˜´ªÂž¦˜µ¤ ˜nœ´ „ª·‹´¥Å¤n°µ‹˜¦ª‹¡Å—oÁœºÉ°Š¤µ‹µ„„µ¦˜´ÊŠ¦³—´
œ´¥Î µ‡´šµŠ™· ˜·š¸É ¦³—´˜É뵂 Ž¹ÉŠÄœµŠ™µœ„µ¦–r‡ª¦Ä®o ‡ªµ¤œÄ‹˜n°„µ¦Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 2
Á¡·É¤ ¹Êœ ¨³¨—‡ªµ¤œÄ‹Äœ„µ¦Á„·—‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 1 œo°¥¨Š š¸É œÎ µ¤µŽ¹ÉŠ°Îµœµ‹„µ¦š—°š¸É
Á®¤µ³¤ (Hays. 1994: 283 – 285) —´Šš¸É ñÁª¨¨r (Howell.1992: 204) „¨nµªªnµ „µ¦°°„Šµœª·‹´¥Á·Šš—¨°Šš¸É
Á o¤Šª—˜n °„µ¦‡ª‡»¤¦³—´œ´¥Îµ‡´šµŠ™·˜· ®¦º°„µ¦š¸É˜´ÊŠ¦³—´œ´¥Îµ‡´šµŠ™·˜·Äœ¦³—´š¸É˜Éε ð„µÄœ„µ¦Á„·—
‡ªµ¤‡¨µ—Á‡¨ºÉ°œž¦³Á£šš¸É 2 ‹³¼ Š ¹ÊœÂ¨³°Îµœµ‹Äœ„µ¦š—° (Power of the test) ‹³¤¸‡nµ¨—¨Š ¨³°Š°µ‹
œ´ ¥¡´•œr (2544: 4 – 5) „¨nµªªnµ „µ¦°°„Šµœª·‹´¥š¸É˜´ÊŠ¦³—´œ´¥Îµ‡´šµŠ™·˜·¦³—´˜Éε šÎµÄ®o¥µ„˜n °„µ¦˜¦ª‹
¡Ÿ¨ °Š„µ¦‹´—„¦³šÎ µšµŠ„µ¦š—¨°Š œ´„ª·‹´¥‡ª¦°°„Šµœª· ‹´¥Ä®o¤¸ °Îµœµ‹„µ¦š—°š¸ÉÁ®¤µ³¤Â¨³
°—‡¨o°Š„´£µ¡„µ¦–r°Š„µ¦šÎµ„µ¦ª· ‹´¥ ¤µ„„ªnµ„µ¦š¸Éœ´ „ª·‹´¥œÄ‹Á¡¸ ¥Š„µ¦°°„Šµœª·‹´¥Á¡ºÉ °Ä®o¤¸
œ´¥Îµ‡´šµŠ™·˜·
ÇÁ±œ (Cohen. 1988) Áž}œœ´„™·˜·‡œÂ¦„š¸ÉÄ®o‡ªµ¤œÄ‹Â¨³Ä®o‡ªµ¤Îµ‡´˜n°°Îµœµ‹„µ¦š—°Äœ„µ¦
ª·‹´¥—oµœ¡§˜· „¦¦¤«µ˜¦r¨³´Š‡¤«µ˜¦r ×¥„¨nµªªnµ °Îµœµ‹„µ¦š—° °Š™·˜·š—° ‡º° ‡ªµ¤œnµ‹³Áž}œš¸É
‹³œÎµÅž¼n„µ¦ž’· Á›¤¤˜·“µœ„¨µŠÁ¤ºÉ °¤¤˜·“µœ„¨µŠÁž}œÁšÈ‹ ¨³Áž}œnªœš¸É ε‡´š¸É »— °Šµ¦œÁš«Á„¸É¥ª„´ 