109 年 國立臺灣大學語言學研究所《語言實驗方法》
第 1 題
Please answer all questions in English.
In 1939, Mary Tudor, a master's student at Department of Psychology, University of Iowa, published her thesis, entitled "An experimental study of the effect of evaluative labeling of speech fluency". Below is an excerpt from her thesis regarding her research questions.
This study was designed to answer the following questions:
- Will removing the label "stutterer" from those who have been so labelled have any effect on their speech fluency?
- Will endorsement of the label "stutterer" previously applied to an individual have any effect on his speech fluency?
- Will endorsement of the label "normal speaker" previously applied to an individual have any effect on his speech fluency?
- Will labelling a person, previously regarded as a normal speaker, a "stutterer" have any effect on his speech fluency?
Her design for the study was shown in the exerpt below.
The children used in this study were taken from the Soldiers and Sailors Orphans' Home in Davenport, Iowa. On January 17, 1939 a survey of 256 children picked at random from the preschool to the ninth grade, inclusive, was made by five individuals trained in speech pathology. To these were added all of the children regarded as stutterers by the teachers and matrons in the institution... From the total group of 256 children 22 were selected for the main part of the present-study...
These 22 children were first divided into two groups. The first group consisted of ten children, who were all of the so-called stutterers in the institution; that is, all of the children who had been labelled stutterers by the teachers and matrons in the institution. The second group consisted of twelve normal speakers with varying degrees of fluency selected at random by the five judges from the 256 children surveyed.
Each of these two groups was then divided. Group IA consisted of five children who had been labelled "stutterers" by members of the institution. An attempt was made to remove the label "stuttering" from the children in this group; that is, they were told that they were not stutterers, but normal speakers who had been erroneously called stutterers. Group IB consisted of the other five children who had been labelled "stutterers" by members of the institution. In the case of these children the judges endorsed the label. Group IIA consisted of six normal speakers with varying degrees of fluency. To this group the judges attached the label "stuttering"; that is, they were told that the type of speech interruptions they were having indicated that they were stutterers. Group IIB consisted of six normal speakers matched in age, sex, intelligence, and fluency with the corresponding six normal speakers in Group IIA. No negative evaluative label was attached to this group. They were merely complimented on their speaking ability, and the label "normal speaker" was endorsed in each case.
The following are excerpts of the statements she made to the four groups of children:
Group IA: ...Many children have this same kind of trouble that you seem to be having. But it really isn't trouble; it's just a certain stage that children to through. In no time you'll outgrow it and you will be able to speak even much better than you are speaking now...
Group IB: No particular instructions or suggestions were given, but a number of questions were asked of each child. Some of the questions were:
- How long have you stuttered?
- When did you begin to stutter?
- What happened that caused you to stutter?
- Was it a person, an incident, or an accident?
- What were you told to do about it?
Group IIA: The staff has come to the conclusion that you have a great deal of trouble with your speech. The type of interruptions which you have are very undesirable... You have many of the symptoms of a child who is beginning to stutter. In fact, you are beginning to stutter...Make up your mind that you are going to speak without a single interruption. It's absolutely necessary that you do this. Do anything to keep from stuttering... Don't ever speak unless you can do it right...
Group IIB: ... You speak very well. Your speech is of very good quality. Speak whenever you have an opportunity. You have the earmarks of a fine speaker.
Tudor also informed the matrons and teachers in the institution about the group of children from which the label "stuttering" had been removed (Group IA) and the group of normal speakers to whom the label "stutterer" had been applied (Group IIA). The statements given were excerpted as follows:
Group IA: The staff has come to the conclusion that these children are not stutterers. They exhibit a special type of speech which they will soon outgrow... They are only going through a temporary stage in speech development which they will soon overcome. Pay no attention whatsoever to their speech. Do not work on it or try to correct it. Take their minds off of it as much as possible. Don't, above all things, ever call it "stuttering", because it isn't stuttering and they will soon outgrow it.
Group IIA: The staff has come to the conclusion that these children show definite symptoms of stuttering. The type of interruptions that they are having very frequently turn into stuttering... You should impress upon them the value of good speech, and that in order to have good speech one has to speak fluently. Watch their speech all of the time very carefully and stop them when they have interruptions; stop them and have them say it over. Don't allow them to speak unless they can say it right. They should be made very conscious of their speech, and also they should be given opportunities to talk so that their mistakes can be pointed out to them.
The five judges gave fluency ratings on a five-point scale for the 22 subjects before and after the experiment. Among them, Tudor was the only one who knew the label that had been attached to each subject during the experimental period. She also recorded the number of speech interruptions these children made. The results could be seen in the following table. The numbers before the slash are fluency ratings and the numbers after are percentages of speech interruptions.
登入後即可作答並保存紀錄。
本題主要在測試考生對實驗設計基本要素的理解,包括實驗設計的類型、自變項(independent variable)、依變項(dependent variable)、因子(factor)及其水準(level)。
實驗設計類型:
此研究採用了因子設計 (Factorial Design),因為它同時操弄了兩個主要因子,並觀察它們對依變項的影響。更具體來說,這是一個2x2 因子設計,因為有兩個二水準因子。
- 因子一:標籤的性質 (Nature of Label)
- 水準一:移除標籤(從「口吃者」移除標籤)
- 水準二:標籤的確認(「口吃者」標籤被確認,或給予「正常說話者」標籤)
- 因子二:標籤的內容 (Content of Label)
- 水準一:負面標籤(「口吃者」)
- 水準二:正面標籤(「正常說話者」)
這兩個因子組合起來,形成了四個實驗組:
- Group IA (移除負面標籤): 原本被標記為「口吃者」,但被告知他們不是口吃者,而是正常的說話者。
- Group IB (確認負面標籤): 原本被標記為「口吃者」,並且此標籤被確認。
- Group IIA (給予負面標籤): 原本是正常說話者,但被標記為「口吃者」。
- Group IIB (確認正面標籤): 原本是正常說話者,被給予「正常說話者」標籤的確認,並受到讚美。
自變項 (Independent Variables, IVs):
自變項是研究者操弄或改變的變數,用來觀察其對依變項的影響。在此研究中,自變項是:
- 標籤的移除或確認 (Label Removal vs. Label Endorsement):這適用於原本被標記為口吃者的兒童。
- 標籤的類型 (Type of Label):這包括「口吃者」標籤(負面)與「正常說話者」標籤(正面)。
更精確地說,我們可以將自變項定義為:
- 標籤的操作 (Label Manipulation):這是由以下兩個因子組合而成:
- 因子 A:標籤的「存在」或「移除」 (Presence/Removal of Label):
- 水準 1: 移除「口吃者」標籤 (Group IA)
- 水準 2: 確認「口吃者」標籤 (Group IB)
- 因子 B:標籤的「性質」 (Nature of Label):
- 水準 1: 負面標籤:「口吃者」(Group IA, IB)
- 水準 2: 正面標籤:「正常說話者」(Group IIB)
- 因子 C:標籤的「給予」或「確認」 (Assignment/Endorsement of Label):
- 水準 1: 給予「口吃者」標籤 (Group IIA)
- 水準 2: 確認「正常說話者」標籤 (Group IIB)
- 因子 A:標籤的「存在」或「移除」 (Presence/Removal of Label):
或者,更簡潔地,我們可以將其視為兩個主要因子:
- 因子 1:標籤的性質 (Nature of Label)
- 水準 1: 負面標籤 (Stutterer)
- 水準 2: 正面標籤 (Normal Speaker)
- 因子 2:標籤的處理方式 (Label Treatment)
- 水準 1: 移除標籤 (Removal) / 轉為正面標籤 (Assignment to Positive)
- 水準 2: 確認標籤 (Endorsement) / 負面標籤 (Assignment to Negative)
從實驗設計的結構來看,我們可以更清晰地理解為:
- Group IA: (原本是 Stutterer) -> Removal of "Stutterer" label. (Effect of removing negative label)
- Group IB: (原本是 Stutterer) -> Endorsement of "Stutterer" label. (Effect of confirming negative label)
- Group IIA: (Normal Speaker) -> Assignment of "Stutterer" label. (Effect of assigning negative label to normal speaker)
- Group IIB: (Normal Speaker) -> Endorsement of "Normal Speaker" label. (Effect of confirming positive label)
因此,自變項可以被理解為「標籤的類型(口吃者 vs. 正常說話者)」以及「標籤的處理方式(移除 vs. 確認 vs. 給予)」。
依變項 (Dependent Variables, DVs):
依變項是研究者測量,以評估自變項影響的變數。在此研究中,依變項是:
- 說話流暢度評分 (Speech Fluency Rating):由五位評審以五點量表測量。
- 說話中斷次數的百分比 (Percentage of Speech Interruptions):由研究者記錄。
因子的種類與水準:
- 因子一:標籤的性質 (Nature of Label)
- 種類:二水準因子 (Two-level factor)
- 水準:負面標籤(「口吃者」)vs. 正面標籤(「正常說話者」)
第 2 題
Please answer the following questions based on the above information:
- Please explain what kind of experimental design this is. Remember to include the
independent and dependent variables and the type and level of factors, among others.
(15%) - What kind of statistical analyses would you use for the data set? Please explain your
choice(s). (20%) - In your opinion, how do the results of the study answer the research questions? Please
substantiate your claims by providing some data analyses (20%) - What are some of the flaws of this study? (20%)
- Do you think the research questions are legitimate? How would you have designed
the experiment instead to make it more sound? (25%)
登入後即可作答並保存紀錄。
本題主要在於評估考生對於統計分析方法的選擇與應用能力,特別是針對實驗數據的處理。考生需要根據研究設計和依變項的性質,提出合適的統計檢定方法,並說明理由。
核心觀念:
實驗數據的分析,特別是比較多組別之間的差異,以及考慮到實驗前後的測量。
資料性質:
- 依變項: 說話流暢度評分 (五點量表) 和說話中斷百分比 (連續變數)。
- 分組: 四個實驗組 (IA, IB, IIA, IIB)。
- 設計: 實驗前後測 (Before/After) 的測量。
統計分析方法建議:
-
針對「說話流暢度評分」 (Speech Fluency Rating):
- 前測與後測的比較 (Within-subjects comparison):
- 對於每一組,可以進行配對樣本 t 檢定 (Paired-samples t-test) 來比較實驗前後的平均流暢度評分是否有顯著差異。
- 組間比較 (Between-subjects comparison):
- 事後檢定 (Post-hoc tests):在進行了整體性的檢定(如 ANOVA)後,用於比較各組別之間後測平均流暢度評分的差異。
- 單因子變異數分析 (One-way ANOVA):用於比較四個實驗組 (IA, IB, IIA, IIB) 在後測 (After) 的平均流暢度評分是否有顯著差異。
- 混合設計變異數分析 (Mixed-design ANOVA):這是最適合此研究設計的方法。
- 自變項 (Within-subjects factor):時間(前測 vs. 後測)。
- 自變項 (Between-subjects factor):實驗組別(IA, IB, IIA, IIB)。
- 依變項:流暢度評分。
- 此方法可以同時檢驗時間的主效果(前後測是否有差異)、組別的主效果(各組平均流暢度是否有差異),以及最重要的時間與組別的交互作用 (Interaction Effect)。交互作用檢驗的意義在於,不同組別的兒童在實驗前後的流暢度變化是否不同,這直接關聯到標籤處理的效果。
- 前測與後測的比較 (Within-subjects comparison):
-
針對「說話中斷百分比」 (Percentage of Speech Interruptions):
- 這是一個連續變數,且有前測與後測數據。
- 前測與後測的比較 (Within-subjects comparison):
- 對於每一組,可以進行配對樣本 t 檢定 (Paired-samples t-test) 來比較實驗前後的平均中斷百分比是否有顯著差異。
- 組間比較 (Between-subjects comparison):
- 單因子變異數分析 (One-way ANOVA):用於比較四個實驗組 (IA, IB, IIA, IIB) 在後測 (After) 的平均中斷百分比是否有顯著差異。
- 混合設計變異數分析 (Mixed-design ANOVA):
- 自變項 (Within-subjects factor):時間(前測 vs. 後測)。
- 自變項 (Between-subjects factor):實驗組別(IA, IB, IIA, IIB)。
- 依變項:中斷百分比。
- 同樣地,此方法能檢驗時間主效果、組別主效果以及最重要的時間與組別的交互作用。
其他可能的分析:
- 共變數分析 (ANCOVA):由於研究者提到了匹配 (matched) 的情況(Group IIB 和 IIA),且有前測數據,可以將前測的數據作為共變數 (Covariate),來分析後測的依變項,以控制前測的差異,進而更精確地檢驗組別的處理效果。
- 例如,使用 ANCOVA 分析後測的中斷百分比,並將前測的中斷百分比作為共變數,以四個組別作為自變項。
- 檢定數據的常態性與變異數同質性:在進行 ANOVA 或 t-test 前,應檢定數據是否符合這些參數檢定的假設。如果不符合,則需要考慮非參數檢定,例如:
第 3 題
Please answer the following questions based on the above information:
- Please explain what kind of experimental design this is. Remember to include the
independent and dependent variables and the type and level of factors, among others.
(15%) - What kind of statstical analyses would you use for the data set? Please explain your
choice(s). (20%) - In your opinion, how do the results of the study answer the research questions? Please
substantiate your claims by providing some data analyses (20%) - What are some of the flaws of this study? (20%)
- Do you think the research questions are legitimate? How would you have designed
the experiment instead to make it more sound? (25%)
登入後即可作答並保存紀錄。
本題要求考生根據提供的數據,分析研究結果如何回答研究問題,並需要進行初步的數據分析來支持論點。這考驗了考生解讀數據、連結研究問題與結果、以及進行簡要數據分析的能力。
核心觀念:
數據解讀、統計結果應用於回答研究問題、初步數據分析。
研究問題回顧:
- 移除「口吃者」標籤對說話流暢度的影響?
- 確認「口吃者」標籤對說話流暢度的影響?
- 確認「正常說話者」標籤對說話流暢度的影響?
- 給予「口吃者」標籤給正常說話者對說話流暢度的影響?
數據表格解讀:
表格提供了 22 位受試者在實驗前 (Before) 和實驗後 (After) 的流暢度評分 (Rating) 和中斷百分比 (Percentage of Interruptions)。
- 流暢度評分 (Rating):分數越高代表越流暢。
- 中斷百分比 (%):百分比越低代表越流暢。
初步數據分析與回答研究問題:
為了回答研究問題,我們需要比較各組別前後測的變化,以及組別間的差異。最直接的方式是計算各組的平均變化量,並進行比較。
計算各組平均值與變化量:
我們需要計算每個組別 (IA, IB, IIA, IIB) 在「流暢度評分」和「中斷百分比」上的平均值,以及實驗前後的平均差異。
| Group | Subject | Before Rating / % | After Rating / % | Rating Change (After - Before) | % Change (After - Before) |
|---|---|---|---|---|---|
| IA | 1 | 2.3/18.37 | 3.2/14.25 | +0.9 / -4.12 | |
| 2 | 1.4/9.99 | 2/8.76 | +0.6 / -1.23 | ||
| 3 | 4.2/8.65 | 3.2/9.84 | -1.0 / +1.19 | ||
| 4 | 3.2/30.97 | 2.2/10.05 | -1.0 / -20.92 | ||
| 5 | 4.0/10.07 | 4.0/9.52 | 0.0 / -0.55 | ||
| IA Avg | 3.02/15.40 | 3.12/10.88 | +0.10 / -6.52 | ||
| IB | 6 | 2.8/1.19 | 3.4/15.24 | +0.6 / +14.05 | |
| 7 | 2.6/11.43 | 3.2/8.41 | +0.6 / -3.02 | ||
| 8 | 2.0/17.81 | 1.8/13.28 | -0.2 / -4.53 | ||
| 9 | 2.1/33.07 | 2.2/27.63 | +0.1 / -5.44 | ||
| 10 | 3.6/3.81 | 3.4/2.42 | -0.2 / -1.39 | ||
| IB Avg | 2.62/14.06 | 2.82/12.00 | +0.20 / -2.06 | ||
| IIA | 11 | 2.6/11.63 | 2.0/14.39 | -0.6 / +2.76 | |
| 12 | 2.8/8.06 | 2.8/10.97 | 0.0 / +2.91 | ||
| 13 | 3.0/7.09 | 3.6/4.83 | +0.6 / -2.26 | ||
| 14 | 2.5/7.96 | 2.5/12.22 | 0.0 / +4.26 | ||
| 15 | 3.0/6.15 | 3.8/8.91 | +0.8 / +2.76 | ||
| 16 | 3.1/6.62 | 2.8/12.47 | -0.3 / +5.85 | ||
| IIA Avg | 2.85/8.17 | 2.85/10.42 | 0.00 / +2.25 | ||
| IIB | 17 | 3.0/10.50 | NA/6.88 | NA | |
| 18 | 2.8/7.62 | 2.6/3.84 | -0.2 / -3.78 | ||
| 19 | 2.5/5.54 | 2.8/2.16 | +0.3 / -3.38 | ||
| 20 | 3.4/4.94 | 3.2/4.61 | -0.2 / -0.33 | ||
| 21 | 2.3/15.53 | 3.0/9.88 | +0.7 / -5.65 |
第 4 題
Please answer the following questions based on the above information:
- Please explain what kind of experimental design this is. Remember to include the
independent and dependent variables and the type and level of factors, among others.
(15%) - What kind of statstical analyses would you use for the data set? Please explain your
choice(s). (20%) - In your opinion, how do the results of the study answer the research questions? Please
substantiate your claims by providing some data analyses (20%) - What are some of the flaws of this study? (20%)
- Do you think the research questions are legitimate? How would you have designed
the experiment instead to make it more sound? (25%)
登入後即可作答並保存紀錄。
本題旨在評估考生對研究方法的批判性思考能力,包括識別研究設計的潛在缺陷,以及提出改進方案。這需要考生具備紮實的實驗設計與統計學知識,並能從多個角度審視研究的嚴謹性。
核心觀念:
研究方法論的批判、實驗設計的限制、改進研究設計。
研究的潛在缺陷 (Flaws of the Study):
-
樣本量過小 (Small Sample Size):
- 每組只有 5-6 名兒童 (IA: 5, IB: 5, IIA: 6, IIB: 6,扣除 NA 的 17 號)。
- 影響: 樣本量小會降低統計檢定的效力 (statistical power),使得即使存在真實效果,也可能無法檢測到顯著差異。同時,結果的可推論性 (generalizability) 受限,難以推論到更廣泛的兒童族群。
-
實驗時間過短 (Short Duration of Intervention):
- 研究描述了實驗設計和數據收集的時間點,但未明確實驗的持續時間。從「Before」到「After」的測量,若間隔時間很短(例如一天或幾天),則效果可能只是暫時性的,無法反映長期標籤效應。
- 影響: 效果可能不持久,對兒童行為的長期改變可能無效。
-
依變項測量的主觀性與客觀性問題:
- 流暢度評分 (Rating):由五位評審進行,雖然有評分標準,但仍可能帶有主觀性,且五點量表本身解析度有限。Subject 17 的 NA 值也顯示評分的潛在問題。
- 中斷百分比 (%):是較客觀的測量,但 OCR 文本中提到「The numbers before the slash are fluency ratings and the numbers after are percentages of speech interruptions」,這意味著數據表格中的數字格式為 "Rating/Percentage"。例如 Subject 1 的 Before 是 2.3/18.37。這表示流暢度評分是 2.3,中斷百分比是 18.37%。
- 影響: 依賴主觀評分可能引入評分者偏誤 (rater bias) 或評分者間信度 (inter-rater reliability) 問題。
-
實驗者偏誤的可能性 (Potential Experimenter Bias):
- 研究提到「Among them, Tudor was the only one who knew the label that had been attached to each subject during the experimental period」。這表示研究者 Tudor 知道每個孩子的標籤。
- 影響: 雖然 Tudor 知道標籤,但她並沒有直接與孩子們互動進行「標籤」的告知,而是由「staff」或「judges」執行。然而,如果 Tudor 在數據分析或觀察過程中,無意中受到標籤知識的影響,就可能產生實驗者偏誤。但這種設計(單盲)在某些情況下是必要的,以保證實驗操弄的準確性。
-
控制組的選擇與對照 (Control Group Selection and Comparison):
- 研究將兒童分成四組,但各組之間的對照關係需要仔細審視。
- IA vs. IB:比較移除負面標籤與確認負面標籤的效果。
- IIA vs. IIB:比較給予負面標籤與確認正面標籤的效果。
- 潛在問題: Group IIB (確認正面標籤) 雖然是正面對照,但其「正常說話者」的身份與 Group IA/IB 的「原口吃者」身份不同。這使得直接比較 IA/IB 與 IIB 的效果變得複雜,因為除了標籤處理,個體原先的狀態也不同。
- 研究將兒童分成四組,但各組之間的對照關係需要仔細審視。
-
「標籤」操作的精確性與一致性:
- 雖然描述了對每個組別的「告知」內容,但實際執行時,不同工作人員(staff, judges)的語氣、態度、執行細節可能存在差異,這會影響標籤操弄的均一性。
- 影響: 標籤操弄的不一致性會增加噪音,降低實驗的內部效度。
-
對「口吃」定義的模糊性:
- 研究採用了「口吃者」和「正常說話者」的標籤,但對於「口吃」的定義,特別是「輕微口吃」或「口吃症狀」的界定,在研究設計中並未非常明確。
- 影響: 這可能導致對「口吃者」的初始分類不夠精確。
-
倫理考量 (Ethical Considerations):
- 將「口吃者」標籤給予原本正常的兒童 (Group IIA) 存在倫理爭議。雖然研究者聲稱這是一種實驗操弄,但可能對兒童造成心理傷害。
改進研究設計的建議:
-
增加樣本量 (Increase Sample Size):
- 招募更多符合條件的兒童,以提高統計效力,增加研究結果的可推論性。
-
延長實驗處理時間 (Extend Intervention Duration):
- 讓標籤處理持續更長的時間(例如數週或數月),以觀察其長期效果,並區分短期和長期影響。
-
標準化依變項測量 (Standardize Dependent Variable Measurement):
- 客觀測量: 盡可能依賴客觀的行為指標,例如使用聲學分析軟體來量化說話中斷的類型、頻率和持續時間。
第 5 題
Please answer the following questions based on the above information:
- Please explain what kind of experimental design this is. Remember to include the
independent and dependent variables and the type and level of factors, among others.
(15%) - What kind of statstical analyses would you use for the data set? Please explain your
choice(s). (20%) - In your opinion, how do the results of the study answer the research questions? Please
substantiate your claims by providing some data analyses (20%) - What are some of the flaws of this study? (20%)
- Do you think the research questions are legitimate? How would you have designed
the experiment instead to make it more sound? (25%)
登入後即可作答並保存紀錄。
本題是綜合題,要求考生評估研究問題的合理性,並提出一個更優化的實驗設計來回答這些問題。這需要考生對語言實驗方法有深入的理解,能夠從理論和實踐上進行創新。
核心觀念:
研究問題的價值與合理性、實驗設計的優化、理論與實務的結合。
研究問題的合理性 (Legitimacy of Research Questions):
研究問題是合理的 (legitimate) 且具有重要意義的。
- 理論意義: 這些問題觸及了「標籤效應」(labeling effect) 或「羅森塔爾效應」(Rosenthal effect) 在語言發展和行為改變中的應用。它探討了社會心理學中的自我實現預言 (self-fulfilling prophecy) 如何影響兒童的語言流暢度。理解標籤如何影響個體表現,對於教育、心理輔導和社會互動都有重要的啟示。
- 實務意義:
- 對於如何對待有語言困難(如口吃)的兒童,研究結果能提供循證的建議。例如,是應該移除負面標籤,還是確認正面標籤?
- 對於如何對待「正常」兒童,研究也探討了負面標籤的潛在危害,以及正面標籤的穩定作用。
- 這對於教師、家長和治療師如何與兒童溝通,如何形成對兒童的期望,具有直接的指導價值。
因此,研究問題本身是值得探究的,並且有潛力產生重要的學術和實務貢獻。
改進實驗設計的建議 (How to Design the Experiment Instead to Make it More Sound):
基於前一題的缺陷分析,以下提出一個更為優化的實驗設計:
研究目標: 探究不同性質和處理方式的語言標籤對兒童說話流暢度的影響。
優化設計理念:
- 明確的實驗因子與水準:將標籤的「性質」(正面/負面)和「處理方式」(賦予/確認/移除/無處理)更系統地進行組合。
- 嚴謹的隨機對照設計:確保各組的隨機分派,並設置更完善的對照組。
- 客觀且多樣化的依變項測量:結合客觀指標和主觀報告。
- 延長實驗處理時間:觀察長期效果。
- 加強倫理考量:避免對兒童造成不必要的傷害。
建議的實驗設計:
1. 受試者 (Participants):
- 招募一定數量的兒童(例如 8-10 歲),他們在實驗前被評估為「口吃」或「正常說話者」(基於標準化的口吃診斷標準,例如 DSM-5 或専門機構的診斷)。
- 樣本量:每組至少 20-30 名兒童,以確保足夠的統計效力。
2. 實驗因子 (Experimental Factors):
- 因子 A:個體原先的狀態 (Original Status)
- 水準 1: 口吃者 (Stutterer)
- 水準 2: 正常說話者 (Normal Speaker)
- 因子 B:標籤處理 (Label Treatment)
- 水準 1: 負面標籤處理 (Negative Label Treatment)
- 子水準 1.1: 確認負面標籤 (Confirm Negative Label) - 如 Tudor 研究的 IB 組。
- 子水準 1.2: 移除負面標籤 (Remove Negative Label) - 如 Tudor 研究的 IA 組。
- 子水準 1.3: 給予負面標籤 (Assign Negative Label) - 如 Tudor 研究的 IIA 組。
- 水準 2: 正面標籤處理 (Positive Label Treatment)
- 子水準 2.1: 確認正面標籤 (Confirm Positive Label) - 如 Tudor 研究的 IIB 組。
- 子水準 2.2: 給予正面標籤 (Assign Positive Label) - 將原本是口吃者但被標記為「正常說話者」的組別。
- 水準 3: 無標籤處理 (No Label Treatment) - 作為對照組,接受與其他組相同的環境,但對其說話不做任何標籤評價。
- 水準 1: 負面標籤處理 (Negative Label Treatment)
3. 實驗設計結構:
這是一個三因子設計,但考慮到因子 A 和 B 的水準組合,可以更精確地描述為:
-
針對「原為口吃者」的受試者 (Factor A = Stutterer):
- 組別 S-N-C (Stutterer - Negative Label - Confirm): 確認「口吃者」標籤。
- 組別 S-N-R (Stutterer - Negative Label - Remove): 移除「口吃者」標籤。
- 組別 S-P-A (Stutterer - Positive Label - Assign): 給予「正常說話者」標籤。
- 組別 S-N-A (Stutterer - Negative Label - Assign): (此組別在 Tudor 研究中未直接體現,可視為探討將「口吃者」標籤給予原本是口吃者但被認為較嚴重的群體,但此設計有倫理風險,可省略或謹慎設計)。
- 組別 S-NL (Stutterer - No Label): 無標籤對照組。
-
針對「原為正常說話者」的受試者 (Factor A = Normal Speaker):
- 組別 N-P-C (Normal Speaker - Positive Label - Confirm): 確認「正常說話者」標籤。
- 組別 N-P-A (Normal Speaker - Positive Label - Assign): 給予「正常說話者」標籤 (與 N-P-C 類似,但強度或方式略有不同)。
- 組別 N-N-A (Normal Speaker - Negative Label - Assign): 給予「口吃者」標籤 (倫理風險較高)。
- 組別 N-NL (Normal Speaker - No Label): 無標籤對照組。
簡化版設計:
為了控制複雜性並聚焦核心問題,可以採用更精簡的設計,例如:
- 受試者群體:
- 兒童診斷為口吃者 (Stutterer)
- 兒童診斷為正常說話者 (Normal Speaker)