Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
外国語教育(研究)における量的データの視覚化と解釈
Search
Ken Urano
August 06, 2019
Education
0
970
外国語教育(研究)における量的データの視覚化と解釈
FLEAT VII (LET2019) ワークショップ
2019/08/06
@早稲田大学
Ken Urano
August 06, 2019
Tweet
Share
More Decks by Ken Urano
See All by Ken Urano
The Task is not the End: The Role of Task Repetition and Sequencing In Language Teaching
uranoken
0
250
学習者を対象にした英語教育研究における倫理的配慮
uranoken
0
740
学習者データを「見る」:外国語教師のためのデータの入力、分析、解釈方法
uranoken
0
890
英語教育研究でエビデンスを「つくる」:メタ分析、再現性、追試
uranoken
0
1.1k
タスク·ベースの英語授業:基本的な考え方とデザイン方法
uranoken
0
1.1k
英語の授業をタスクで組み立てる
uranoken
0
1.2k
Designing Task-based ESP Syllabi: Two Cases from an English for Business Purposes Program
uranoken
0
1.2k
第二言語習得と外国語教育における 「文法知識」の位置づけ
uranoken
0
1.2k
英語教育研究の始め方・進め方:目的に合致した手法選択の重要性
uranoken
1
890
Other Decks in Education
See All in Education
Stratégie de marketing digital - les fondamentaux
martine
0
160
LinkedIn
matleenalaakso
0
3.4k
1030
cbtlibrary
0
330
AWS Well-Architected Labを活用してつよつよAWSエンジニアになろう!!! #jawsug_tokyo
masakiokuda
0
260
Repaso electricidade e electrónica
irocho
0
220
Future Trends and Review - Lecture 12 - Web Technologies (1019888BNR)
signer
PRO
0
2.6k
Da Necessidade da Devoção à Virgem Santíssima
cm_manaus
0
100
CSS3 and Responsive Web Design - Lecture 5 - Web Technologies (1019888BNR)
signer
PRO
1
2.5k
Medidas en informática
irocho
0
480
Sähköiset kyselyt, kokeet ja arviointi
matleenalaakso
1
17k
Medicare 101 for 2025
robinlee
PRO
0
380
自己紹介 / who-am-i
yasulab
PRO
2
4.4k
Featured
See All Featured
How to Think Like a Performance Engineer
csswizardry
22
1.3k
No one is an island. Learnings from fostering a developers community.
thoeni
19
3.1k
How STYLIGHT went responsive
nonsquared
96
5.3k
A Tale of Four Properties
chriscoyier
157
23k
The Illustrated Children's Guide to Kubernetes
chrisshort
48
49k
Testing 201, or: Great Expectations
jmmastey
41
7.2k
Thoughts on Productivity
jonyablonski
68
4.4k
The Language of Interfaces
destraynor
155
24k
Why Our Code Smells
bkeepers
PRO
335
57k
Raft: Consensus for Rubyists
vanstee
137
6.7k
Faster Mobile Websites
deanohume
305
30k
Distributed Sagas: A Protocol for Coordinating Microservices
caitiem20
330
21k
Transcript
֎ࠃޠڭҭʢݚڀʣʹ͓͚Δ ྔతσʔλͷࢹ֮Խͱղऍ Ӝ ݚʢւֶԂେֶʣ email:
[email protected]
FLEAT VII / LET2019
@ Waseda University ɹɹ2019. 8. 6. https://www.urano-ken.com/research/let2019
ຊͷࢿྉ
֎ࠃޠڭҭʹܞΘΔࢲͨͪɺݚڀʹ͓͍͚ͯͩͰͳ ͘ɺςετॲཧͱ͍ͬͨ໘Ͱ͝Ζ͔ΒྔԽ ͞ΕͨσʔλΛѻ͍ͬͯ·͢ɻຊϫʔΫγϣοϓͰɺ ڭҭݚڀͰྔతσʔλΛѻ͏ࡍʹ·ͣߦ͏͖σʔλ ͷࢹ֮ԽͱɺσʔλͷಛΛཧղ͢ΔͨΊͷجຊతͳ֓ ೦ͱͯ͠ͷදɾɾޮՌྔͷҙຯʹֶ͍ͭͯͼɺ ϑϦʔͰΦʔϓϯιʔεͷ౷ܭιϑτ jamovi Λͬͯɺ ࣮ࡍʹσʔλͷ؆୯ͳੳ͕Ͱ͖ΔΑ͏ʹͳΔ͜ͱΛ
ࢦ͠·͢ɻ ཁࢫ
ՍۭͷσʔλΛ ༻ҙ͠·ͨ͠
Name* Test A খ ರ 70 Տ େޒ 38 খਿ
Ꮺ 58 ௶Ҫ ج༞ 48 ӬҪ ج༞ 28 ڮޱ ๏࢚ 54 ݪ ཽ 58 ༎ 38 ౻ా ࢰಐ 42 ຊؒ խ 47 ٶ࡚ ৎ༤ 78 ଜҪ 68 ࢁ࡚ ଠ 40 ԣҪ ޛࢤ 50 ґా ༸հ 68 एࢁ ప 57 ༗അ Ղ೫ 64 ઘ ګࢠ 76 ؠҪ ඒՂ 43 ߐ ༝Ӊ 90 ਆ୩ ࣿق 58 ઍՂࢠ 38 ࡔా Ѫࡊ 38 ਿా ඒՂ 43 ⁋ຊ ᜫ 58 ୩ ே߳ 60 Ӭ ͘ΔΈ 48 দ ಹಸ 45 ଜҪ ݁ࢠ 24 ए௬ ·Έ 36 *ʮͳΜͪΌͬͯݸਓใʯͰੜ http://kazina.com/dummy/
Group A Test A খ ರ 70 Տ େޒ 38
খਿ Ꮺ 58 ௶Ҫ ج༞ 48 ӬҪ ج༞ 28 ڮޱ ๏࢚ 54 ݪ ཽ 58 ༎ 38 ౻ా ࢰಐ 42 ຊؒ խ 47 ٶ࡚ ৎ༤ 78 ଜҪ 68 ࢁ࡚ ଠ 40 ԣҪ ޛࢤ 50 ґా ༸հ 68 एࢁ ప 57 ༗അ Ղ೫ 64 ઘ ګࢠ 76 ؠҪ ඒՂ 43 ߐ ༝Ӊ 90 ਆ୩ ࣿق 58 ઍՂࢠ 38 ࡔా Ѫࡊ 38 ਿా ඒՂ 43 ⁋ຊ ᜫ 58 ୩ ே߳ 60 Ӭ ͘ΔΈ 48 দ ಹಸ 45 ଜҪ ݁ࢠ 24 ए௬ ·Έ 36 Group B Test A ؠӬ 52 ২ ҭೋ 59 ย ཽ 61 ࡔݩ ᠳଠ 76 ౡଜ ༏ 45 ా ར 68 ࢙ 63 দҪ Ұಙ 69 ࡾݪ ༟࣍ 43 क ཽ࣍ 51 ੨ Έ͋ 36 ୩ ༏ 51 ؠ୩ ౧ࢠ 39 ্ݪ ܠࢠ 71 ߐޱ Ί͙Έ 26 ٴ ͳͭΈ 79 େ௩ ·͞Έ 55 Ԭ ࿏ࢠ 61 ֯ా ౧ࢠ 89 ݁ҥ 51 ਆށ ࡊʑඒ 71 ֎ࢁ Έ͋ 63 রҪ Έ͖ 41 ࠜ؛ ༏ 41 ࠜ؛ ྱࢠ 83 Ӌా ѥر 93 ࢜ ΈΏ͖ 47 ࢪ ༑߳ 37 ଜా จੈ 52 ٢Ӭ ܙས߳ 41
Group A Test A খ ರ 70 Տ େޒ 38
খਿ Ꮺ 58 ௶Ҫ ج༞ 48 ӬҪ ج༞ 28 ڮޱ ๏࢚ 54 ݪ ཽ 58 ༎ 38 ౻ా ࢰಐ 42 ຊؒ խ 47 ٶ࡚ ৎ༤ 78 ଜҪ 68 ࢁ࡚ ଠ 40 ԣҪ ޛࢤ 50 ґా ༸հ 68 एࢁ ప 57 ༗അ Ղ೫ 64 ઘ ګࢠ 76 ؠҪ ඒՂ 43 ߐ ༝Ӊ 90 ਆ୩ ࣿق 58 ઍՂࢠ 38 ࡔా Ѫࡊ 38 ਿా ඒՂ 43 ⁋ຊ ᜫ 58 ୩ ே߳ 60 Ӭ ͘ΔΈ 48 দ ಹಸ 45 ଜҪ ݁ࢠ 24 ए௬ ·Έ 36 Group B Test A ؠӬ 52 ২ ҭೋ 59 ย ཽ 61 ࡔݩ ᠳଠ 76 ౡଜ ༏ 45 ా ར 68 ࢙ 63 দҪ Ұಙ 69 ࡾݪ ༟࣍ 43 क ཽ࣍ 51 ੨ Έ͋ 36 ୩ ༏ 51 ؠ୩ ౧ࢠ 39 ্ݪ ܠࢠ 71 ߐޱ Ί͙Έ 26 ٴ ͳͭΈ 79 େ௩ ·͞Έ 55 Ԭ ࿏ࢠ 61 ֯ా ౧ࢠ 89 ݁ҥ 51 ਆށ ࡊʑඒ 71 ֎ࢁ Έ͋ 63 রҪ Έ͖ 41 ࠜ؛ ༏ 41 ࠜ؛ ྱࢠ 83 Ӌా ѥر 93 ࢜ ΈΏ͖ 47 ࢪ ༑߳ 37 ଜా จੈ 52 ٢Ӭ ܙས߳ 41 ൺͯΈΑ͏ How?
ᶃ ਤʹͯ͠ΈΑ͏
ώετάϥϜ (Histogram) B A 20 40 60 80 100 Score
Group
๘܈ਤ (Beeswarm) 20 40 60 80 A B Group Score
ശͻ͛ਤ (Box Plot) 20 40 60 80 A B Group
Score
ϰΝΠΦϦϯਤ (Violin Plot) 20 40 60 80 A B Group
Score
֬ີ (Density) B A 30 60 90 Score Group
֬ີ (Density) B A 30 60 90 Score Group
ਤʹͯ͠ΈΑ͏ • ऩूͨ͠σʔλʹͲͷΑ͏ͳಛ͕͋Δ͔ɺ ͬ͘͟ΓѲ͢Δ͜ͱ͕Ͱ͖Δɻ • ͰݟΔ͚ͩͳͷͰɺݫີͳൺֱੳʹ ద͞ͳ͍ɻ
ᶄ ཁͯ͠ΈΑ͏
σʔλͷத৺ͱ Β͖ͭ σʔλͷத৺
ฏۉ ͯ͢ͷσʔλͷ߹ܭΛσʔλͷݸͰ ׂͬͨͷ தԝ ͯ͢ͷσʔλΛখ͍͞ॱʢ·ͨେ͖͍ ॱʣʹฒͨͱ͖ɺਅΜதʹདྷΔ ࠷ස ͯ͢ͷσʔλͷதͰग़ݱճ͕࠷ଟ͍ σʔλͷத৺
Group A Group B ฏۉ 52.1 57.1 தԝ 49.0 53.5
࠷ස 38, 58 41, 51 σʔλͷத৺
ඪ४ภࠩ σʔλͷΒ͖ͭ
• ݸʑͷͱฏۉͱͷࠩΛ̎͠ɺ ͦͷ߹ܭΛσʔλͷͰׂͬͨͷͷฏํࠜ Group A Test A খ ರ 70
Տ େޒ 38 খਿ Ꮺ 58 ௶Ҫ ج༞ 48 ӬҪ ج༞ 28 ڮޱ ๏࢚ 54 ݪ ཽ 58 ༎ 38 ౻ా ࢰಐ 42 (70–52.1)2 = 320.4 (38–52.1)2 = 198.8 (58–52.1)2 = 034.8 . . . ߹ܭ 6828.7 / 30 = 227.6 √ 227.6 = 15.1 Group A ฏۉ 52.1 ←ʢࢄʣ ඪ४ภࠩ
• ݸʑͷͱฏۉͱͷࠩΛ̎͠ɺ ͦͷ߹ܭΛσʔλͷͰׂͬͨͷͷฏํࠜ Group A Test A খ ರ 70
Տ େޒ 38 খਿ Ꮺ 58 ௶Ҫ ج༞ 48 ӬҪ ج༞ 28 ڮޱ ๏࢚ 54 ݪ ཽ 58 ༎ 38 ౻ా ࢰಐ 42 (70–52.1)2 = 320.4 (38–52.1)2 = 198.8 (58–52.1)2 = 034.8 . . . ߹ܭ 6828.7 / 30 = 227.6 √ 227.6 = 15.1 Group A ฏۉ 52.1 ඪ४ภࠩ 15.1 ←ʢࢄʣ ඪ४ภࠩ
0 20 40 60 80 100 0.00 0.01 0.02 0.03
0.04 0 20 40 60 80 100 0.00 0.01 0.02 0.03 0.04 ฏۉ = 50 ͷ߹ ඪ४ภࠩ = 10 ඪ४ภࠩ = 20 34.1% 13.6% 34.1% 34.1% 13.6% 34.1% 13.6% 13.6% ඪ४ภࠩ
0 20 40 60 80 100 0.00 0.01 0.02 0.03
0.04 0 20 40 60 80 100 0.00 0.01 0.02 0.03 0.04 ฏۉ = 50 ͷ߹ ඪ४ภࠩ = 10 ඪ४ภࠩ = 20 ඪ४ภࠩ
ʢ٢ా, 1998, p. 173ʣ ඪ४ภࠩ
ʢ٢ా, 1998, p. 173ʣ ࠩಉ͡ ඪ४ภࠩ
ॏͳΓͷྔ͕ҧ͏ ඪ४ภࠩ
Group A Group B ฏۉ 52.1 57.1 ඪ४ภࠩ 15.1 16.4
Group A Group B 0 20 40 60 80 100
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0 20 40 60 80 100 0.000 0.005 0.010 0.015 0.020 0.025 0.030
Group A Group B 0 20 40 60 80 100
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0 20 40 60 80 100 0.000 0.005 0.010 0.015 0.020 0.025 0.030
ฏۉͷࠩ Group A Group B Group A Group B ฏۉ
52.1 57.1 ඪ४ภࠩ 15.1 16.4 0 20 40 60 80 100 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0 20 40 60 80 100 0.000 0.005 0.010 0.015 0.020 0.025 0.030 ͷҧ͍
ݴ͑ͦ͏ͳ͜ͱ • ฏۉͷൺֱ͚ͩͰෆे • σʔλͷʢΒ͖ͭʣ߹Θͤͯݕ౼ • ͷॏͳΓ͕গͳ͍ํ͕͕ࠩେ͖͍
͏ҰൺͯΈΑ͏ Group A Group B 0 20 40 60 80
100 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0 20 40 60 80 100 0.000 0.005 0.010 0.015 0.020 0.025 0.030
͏ҰൺͯΈΑ͏ 0" 1" 2" 3" 4" 5" 6" 7" 8"
9" 0,10" 11,20" 21,30" 31,40" 41,50" 51,60" 61,70" 71,80" 81,90" 91,100" Group"A" Group"B" ࣮ࡍͷΛϓϩοτͨ͠ͷ
͏ҰൺͯΈΑ͏ 0" 1" 2" 3" 4" 5" 6" 7" 8"
9" 0,10" 11,20" 21,30" 31,40" 41,50" 51,60" 61,70" 71,80" 81,90" 91,100" Group"A" Group"B" ͜ͷॏͳΓେ͖͍ͷʁখ͍͞ͷʁ
ࢦඪ͕΄͍͠
ޮՌྔʢEffect Sizeʣ • ޮՌͷେ͖͞Λ͋ΒΘ͢౷ܭతͳࢦඪ ʢେٱอɾԬా, 2012, p. 44ʣ
ޮՌྔͷछྨ
• ࠩͷେ͖͞Λද͢ࢦඪʢd ʣ • ؔͷڧ͞Λද͢ࢦඪʢr ʣ େ͖͚ͯ̎ͭ͘
ࠩͷେ͖͞Λද͢ࢦඪ Cohen’s d
pooled SD X X d 2 1 − = ←ɹฏۉͷࠩ
←ɹඪ४ภࠩ Cohen’s d ʮ̎ͭͷάϧʔϓͷࠩඪ४ภࠩԿݸʯ
pooled SD X X d 2 1 − = |
52.1 - 57.1| = (15.1 + 16.4) / 2* *ඪຊαΠζ͕ҟͳΔͱ͖ɺSDpooled ͷܭࢉ͏গ͠ෳࡶʹͳΓ·͢ Group A Group B ฏۉ 52.1 57.1 ඪ४ภࠩ 15.1 16.4 Cohen’s d
pooled SD X X d 2 1 − = 5.0
= 15.75 *ඪຊαΠζ͕ҟͳΔͱ͖ɺSDpooled ͷܭࢉ͏গ͠ෳࡶʹͳΓ·͢ Group A Group B ฏۉ 52.1 57.1 ඪ४ภࠩ 15.1 16.4 = 0.32 Cohen’s d
d 0 0.1 0.2 0.3 0.4 0.5 0.6 ॏͳΓ ʢˋʣ
100 92.3 85.7 78.7 72.6 67 61.8 d 0.7 0.8 0.9 1.0 1.1 1.2 1.3 ॏͳΓ ʢˋʣ 57 52.6 48.4 44.6 41.1 37.8 34.7 ޮՌྔ d ͱͷॏͳΓ
0" 1" 2" 3" 4" 5" 6" 7" 8" 9"
0,10" 11,20" 21,30" 31,40" 41,50" 51,60" 61,70" 71,80" 81,90" 91,100" Group"A" Group"B" d = 0.32 ͳͷͰॏͳΓ 3/4 ͙Β͍ ࠶ͼ͜ͷάϥϑ
ͭ·Γ
Group A ͱ Group B ɺ ฏۉʹ 5 ͕ࠩ͋Δ͕ɺ શମͷ
3/4 ॏͳ͍ͬͯΔɻ
ޮՌྔ d ͱॏͳΓͷؔ
pooled SD X X d 2 1 − = ←ɹখ͍͞ํ͕ྑ͍
←ɹେ͖͍ํ͕ྑ͍ d ͕େ͖͘ͳΔʹ ʮฏۉͷ͕ࠩେ͖͘ɺඪ४ภ͕ࠩখ͘͞ͳΔ ͱɺޮՌྔେ͖͘ͳΔɻʯ
• Cohen (1988) • small: d = 0.2, overlap: 85.7%
• e.g., 15ࡀͱ16ࡀͷঁࢠͷࠩ • medium: d = 0.5, overlap: 67.0% • e.g., 14ࡀͱ18ࡀͷঁࢠͷࠩ • large: d = 0.8, overlap: 52.6% • e.g., େֶ৽ೖੜͱPhDऔಘऀͷIQࠩ ޮՌྔͷղऍ
• Plonsky & Oswald (2014) • “L2 field-specific benchmarks” ܈ؒൺֱ
܈ൺֱ small d = 0.40 d = 0.60 medium d = 0.70 d = 1.00 large d = 1.00 d = 1.40 ޮՌྔͷղऍ
ͨͩ͠
• ͜ͷΑ͏ͳࢦඪ͋͘·Ͱ҆ • ࣮ࡍͷղऍݚڀऀࣗͷͰ
༗ҙੑݕఆ ॏͳΓͷେ͖͞Θ͔͚ͬͨͲɺ ͜ͷࠩۮવʁ
؍͞Ε͕ͨࠩۮવੜͨ͡ͷͰ͋Δ Մೳੑʢ֬ʣ ༗ҙੑݕఆ
ʮʢ౷ܭతʣ༗ҙੑʯͱ • ͷલͷσʔλʢඪຊʣ͔ΒΑΓେ͖ͳจ຺ ʢूஂʣΛਪఆ͢Δ • ඪຊͰ؍͞ΕΔࠩɾ͕ؔɺूஂ͔Βͷ ඪຊநग़࣌ͷޡࠩͰੜ͡Δ֬ʢp ʣΛ ܭࢉ͢Δ •
p ͕ج४ʢྟքʣҎԼͰ͋Εʮ༗ҙʯ Ͱ͋ΔʢޡࠩͰͳ͍ʣͱஅ͢Δ
ूஂ ඪɹຊ ਪఆ σʔλղੳ Σ, F, t, p... ूஂͱඪຊ
• ͋ΔඪຊͰಘΒΕͨදʢe.g., ฏۉʣ ͱूஂͷදͱͷࠩ ඪຊޡࠩ
ूஂ μ = 15.3 ඪຊA M = 14.7 ඪຊB M
= 15.9 ඪຊC M = 15.2 ඪຊD M = 15.4 ඪຊE M = 15.1
ूஂ μ = 14.7 ඪຊA M = 14.7 ࣮ࡍ M
= μ ͱͯ͠ਪఆ
• ඪຊͷαΠζ͕େ͖͚Εେ͖͍΄Ͳɺ ඪຊޡࠩখ͘͞ͳΔ • ͭ·Γਪఆͷਫ਼͕ߴ͘ͳΔ ඪຊޡࠩ
t ݕఆ ← ฏۉͷࠩ ← ඪ४ภࠩ2ͷ 1 2 2 2
1 2 1 − + − = n SD SD X X t ↑ ʢ֤܈ͷඪຊαΠζʣ ʢඪຊαΠζ͕͍͠߹ʣ ʢ٢ా, 1998, p. 186ʣ
͜Ε͖ͬ͞ݟͨʁ
pooled SD X X d 2 1 − = ←ɹฏۉͷࠩ
←ɹඪ४ภࠩ Cohen’s d ʮ͜Εʹ n Λ͢ͱ t ͬΆ͍ʂʯ
pooled SD X X d 2 1 − = 1
2 2 2 1 2 1 − + − = n SD SD X X t ʮt ɺޮՌྔʹඪຊαΠζΛՃຯͨ͠ͷʯ
←ɹখ͍͞ํ͕ྑ͍ ←ɹେ͖͍ํ͕ྑ͍ t ͕େ͖͘ͳΔʹ 1 2 2 2 1 2
1 − + − = n SD SD X X t ↑ɹେ͖͍ํ͕ྑ͍
ࣗ༝** 3 4 5 10 20 30 ྟք ྆ଆݕఆ5% 3.182
2.776 2.571 2.228 2.086 2.042 ࣗ༝ 40 50 100 200 500 1,000 ྟք ྆ଆݕఆ5% 2.021 2.009 1.984 1.972 1.965 1.962 *͜ΕΑΓେ͖͍ͩͬͨΒۮવͰͳ͍ͱΈͳ͢ **n1 +n2 -2 t ͷྟք*
1 2 2 2 1 2 1 − + −
= n SD SD X X t *2܈Ͱ n ͕ҟͳΔͱ͖ͷܭࢉ ͏গ͠ෳࡶʹͳΓ·͢ * | 52.1-57.1| = √(15.12 + 16.42) / (30 - 1) ܭࢉͯ͠ΈΑ͏ Group A Group B ฏۉ 52.1 57.1 ඪ४ภࠩ 15.1 16.4
1 2 2 2 1 2 1 − + −
= n SD SD X X t * 5 = 4.14 ܭࢉͯ͠ΈΑ͏ = 1.21 Group A Group B ฏۉ 52.1 57.1 ඪ४ภࠩ 15.1 16.4 *2܈Ͱ n ͕ҟͳΔͱ͖ͷܭࢉ ͏গ͠ෳࡶʹͳΓ·͢
ࣗ༝** 3 4 5 10 20 30 ྟք ྆ଆݕఆ5% 3.182
2.776 2.571 2.228 2.086 2.042 ࣗ༝ 40 50 100 200 500 1,000 ྟք ྆ଆݕఆ5% 2.021 2.009 1.984 1.972 1.965 1.962 t ͷྟք t (58) = 1.21 ༗ҙͰͳ͍
͜͜·Ͱͷ·ͱΊ
• ޮՌྔ Cohen’s d • 2ͭͷάϧʔϓؒͷࠩΛඪ४Խͨ͠ͷ • t ݕఆ •
ޮՌྔʹඪຊޡࠩͷӨڹΛՃຯͯ͠ɺͦͷ͕ࠩ ۮવ؍͞ΕΔ֬Λࣔͨ͠ͷ • ݕఆ౷ܭྔ = ޮՌͷେ͖͞ x ඪຊͷେ͖͞ ʢೆ෩ݪ, 2002, p. 163ʣ
• Cohen’s d ͷؒ: • Hedges’ g • ʹूஂͷඪ४ภࠩʢෆภࢄʹ جͮ͘ඪ४ภࠩʣΛ͏
• Glass’ ⊿ • ʹ౷੍܈ͷඪ४ภࠩΛ͏
ؔͷڧ͞Λද͢ࢦඪ Pearson’s r / r2
• มؒͷؔͷେ͖͞Λද͢ • ࠷େ: 1.0ʢઈରʣ • ࠷খ: 0 • ϐΞιϯͷੵ૬ؔ
r • r2 ʢࢄઆ໌ʣ Pearson’s r / r2
ࢄੳͷ߹ ௐ͍ͨཁҼͷࢄ η2 = ૯ࢄ SSA = SSTotal
ҰཁҼࢄੳ SS df MS F p η2 A (Class)
848 2 424 0.955 .389 .022 Error (Residuals) 37260 84 444 ɾਫຊ (2014) ୈ6ষͷσʔλΛͬͯ jamovi Ͱܭࢉ
ҰཁҼࢄੳ SS df MS F p η2 A (Class)
848 2 424 0.955 .389 .022 Error (Residuals) 37260 84 444 / = / = MS = SS / df
ҰཁҼࢄੳ SS df MS F p η2 A (Class)
848 2 424 0.955 .389 .022 Error (Residuals) 37260 84 444 / = ↑ɹඪຊαΠζ͕େ͖͍ͱ F ͕େ͖͘ͳΔ F = MSA / MSError = 424 / 444 = 0.955
ҰཁҼࢄੳ SS df MS F p η2 A (Class)
848 2 424 0.955 .389 .022 Error (Residuals) 37260 84 444 + η2 = SSA / SSTotal = 848 / 38108 = .022 SSTotal = SSA + SSError = 848 + 37260 = 38108
ޮՌྔͷղऍ
• small: η2 = .01 • medium: η2 = .06
• large: η2 = .14 ਫຊɾ (2008) • ͜ͷΑ͏ͳࢦඪ͋͘·Ͱ҆ • ࣮ࡍͷղऍݚڀऀࣗͷͰ
͜͜·Ͱͷ·ͱΊ
• r ͷޮՌྔ • มؒͷؔͷڧ͞ΛͰࣔͨ͠ͷ • ࠷େͰ 1.0ɺ࠷খͰ 0 •
ࢄੳͰ͏ η2 r2 ͱࣅͨײ͡ • F ͱ η2 ͷҧ͍ඪຊαΠζΛߟྀ͢Δ͔Ͳ͏͔ • ݕఆ౷ܭྔ = ޮՌͷେ͖͞ x ඪຊͷେ͖͞ ʢೆ෩ݪ, 2002, p. 163ʣ
• η2 ͷؒ: • partial η2 • ʹ SSA +
SSError Λ͏ • ω2 • ࢄਪఆͷͨΊͷόΠΞεΛऔΓআ ͍ͨͷ
࣮ࡍʹ ܭࢉͯ͠Έ·͠ΐ͏
• Φʔϓϯιʔεͷ౷ܭϓϩάϥϛϯάݴޠ ɹɹΛ͍͍͢ܗʹͨ͠ιϑτΣΞɻ • GUIͷͨΊײతʹ͑Δɻ • ΦʔϓϯιʔεͰແྉͰ͑Δɻ
https://www.jamovi.org
ϋϯζΦϯ
• Must-read: • Navarro, D. J., & Foxcroft, D. R.
(2019). Learning statistics with jamovi: A tutorial for psychology students and other beginners. (Version 0.70). DOI: 10.24384/hgc3-7p15 • ຊޠ༁͋Γ·͢: • ࣳా࢘༁. jamoviͰֶͿ৺ཧ౷ܭ. https://bookdown.org/sbtseiji/lswjamoviJ/
ҙ
None
• ϑΝΠϧಡΈࠐΈ࣌ʹࣗಈత ʹஅ͞ΕΔมͷछྨ͕ؒ ҧ͍ͬͯΔ͜ͱ͕͋Δɻ • Continuous ࿈ଓม • Ordinal ॱংม
• Nominal ໊ٛม
http://www.langtest.jp Effect Size Calculator @
• σʔλੳ͕ͳΜͰ Ͱ͖ͪΌ͏ͷ͍͢͝ ΣϒΞϓϦ • ޮՌྔ d, g ͱͦͷ৴པ۠ ؒΛܭࢉͯ͘͠ΕΔ
• ਫຊಞ͞Μʢؔେֶʣ ͕։ൃ͠ɺແྉͰެ։ • ͓ྱϏʔϧ·ͨ നϫΠϯͰ
None
None
. ਪଌ౷ܭʢ༗ҙੑݕఆʣͰ͏ ඪ४ภࠩʢSDʣෆภࢄʹجͮ͘ ͷɻn ͷΘΓʹ n–1 Λܭࢉʹ ͍·͢ɻ
None
ࢀߟจݙ • ӳޠڭࢣͷͨΊͷڭҭσʔλ ੳೖ • ༗ҙੑݕఆͷ͘͠Έͦͷݶ քʹ͍ͭͯղઆ
ࢀߟจݙ • ຊʹΘ͔Γ͍͘͢͢͝େ ͳ͜ͱ͕ॻ͍ͯΔ͘͝ॳา ͷ౷ܭͷຊ • େͳ͜ͱΛࣜΛަ͑ͯஸ ೡʹղઆ
ࢀߟจݙ • ֎ࠃޠڭҭݚڀϋϯυϒοΫ • هड़౷ܭɺਪଌ౷ܭɺޮՌྔ ؚΊͯཏతͳҰ
ࢀߟจݙ • ͑ΔͨΊͷ৺ཧ౷ܭ • ޮՌྔʹ͍ͭͯษڧ͢ΔͳΒ ඞಡ
ࢀߟจݙ • ͡Ίͯͷӳޠڭҭݚڀ • ݚڀͷೖޱΛղઆ͢ΔҰɻ ࠓͷ༰ୈ6ষΛิ͢ Δͷ
1. σʔλͷࢹ֮Խʢਤࣔʣ 2. σʔλͷཁʢத৺ͱΒ͖ͭʣ 3. ޮՌྔ • ࠩͷେ͖͞Λද͢ d
• ؔͷڧ͞Λද͢ r 4. ༗ҙੑݕఆʢਪଌ౷ܭʣ 5. jamovi ͱ langtest.jp Ken Urano
[email protected]
https://www.urano-ken.com/research/let2019 ֎ࠃޠڭҭʢݚڀʣʹ͓͚Δ ྔతσʔλͷࢹ֮Խͱղऍ
ࢀߟจݙ • Cohen, J. (1988). Statistical power analysis for the
behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates. • ೆ෩ݪே. (2002). ʰ৺ཧ౷ܭֶͷجૅ: ౷߹తཧղͷͨΊʹʱ౦ژ: ༗൹ֳ. • લాܒ࿕ɾࢁޫཅ (ฤ). (2004). ʰӳޠڭࢣͷͨΊͷڭҭσʔλੳೖ: तۀ͕มΘ ΔςετɾධՁɾݚڀʱ౦ژ: େमؗॻళ. • ਫຊಞɾཧ. (2008). ʮݚڀจʹ͓͚ΔޮՌྔͷใࠂͷͨΊʹ: جૅత֓೦ͱҙ ʯʰӳޠڭҭݚڀʱୈ31߸, 57–66. http://www.mizumot.com/files/ EffectSize_KELES31.pdf • Navarro, D. J., & Foxcroft, D. R. (2019). Learning statistics with jamovi: A tutorial for psychology students and other beginners. (Version 0.70). doi: 10.24384/hgc3-7p15 ʢࣳా࢘༁. jamoviͰֶͿ৺ཧ౷ܭ. https://bookdown.org/sbtseiji/lswjamoviJ/ʣ • େٱอ֗ѥɾԬాݠհ. (2012). ʰ͑ΔͨΊͷ৺ཧ౷ܭ: ޮՌྔɾ৴པ۠ؒɾݕఆྗʱ ౦ژ: Ⴛॻ. • Plonsly, L., & Oswald, F. (2014). How big is “big”? Interpreting effect sizes in L2 research. Language Learning, 64, 878–912. doi: 10.1111/lang.12079 • ཧɾਫຊಞ (ฤ). (2014). ʰ֎ࠃޠڭҭݚڀϋϯυϒοΫ: ݚڀख๏ͷΑΓྑ͍ཧղ ͷͨΊʹ (վగ൛)ʱ౦ژ: দദࣾ. • ӜݚɾཧཅҰɾాதɾ౻ాɾ∁ѥرࢠɾञҪӳथ. (2016). ʰ͡Ίͯͷ ӳޠڭҭݚڀ: ԡ͓͖͍͑ͯͨ͞ίπͱϙΠϯτʱ౦ژ: ݚڀࣾ. • ٢ాण. (1998). ʰຊʹΘ͔Γ͍͘͢͢͝େͳ͜ͱ͕ॻ͍ͯ͋Δ͘͝ॳาͷ౷ܭ ͷຊʱژ: େ࿏ॻ.