Scribd
Upload
8 views

ch6

Uploaded by

黃承瀚106三勇31
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as RTF, PDF, TXT or read online on Scribd
8 views

ch6

Uploaded by

黃承瀚106三勇31
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as RTF, PDF, TXT or read online on Scribd
You are on page 1/ 5

第六章、 質化與量化的研究設計

一、研究的質化與量化取向
量化 質化
實證主義的 詮釋的批判的
硬性資料 軟性資料
重建的邏輯 實務的邏輯
技術體制觀點 超驗觀點
線性研究路線 非線性
變項與假設的語言 個案與脈絡的語言
（精準測量變項）
（考驗因果的假設）

Quantitative Designs 量化設計的特色


• Quantitative social work researchers often refer to their observations as ‘hard data’;
that is, statistical tabulations of occurrences or events.
• 硬性資料
• Quantitative social work researchers are more likely to use reconstructed logic—
highly organized formal language.
• 重建邏輯：研究的邏輯有系統、組織化、共享的明確的步驟與方法
• Quantitative social work researchers use technocratic logic, engaging statistical design
issues.
• 技術體制：研究者是專家（技術人員）、研究問題來自贊助者；研究目的在發現律則以增進效率（滿足官僚體制需求）
• Quantitative social work researchers usually follow a linear path using variables and
hypotheses, emphasizing precise measures, and linking findings to general causal
explanations.
• 線性路徑

Qualitative Designs 質化設計的特色


• Qualitative social work researchers are more likely to refer to their observations as
‘soft data’, these derive from their impressions, statements, photos, maps, or symbols.
• 軟性資料
• Qualitative social work researchers are more likely to use logic in practice, making
informal judgment calls.
• 倚賴實務邏輯進行研究：既定規則少、共享的判準或規範，傳承的非正式智慧
• Qualitative social work researchers use transcendent logic about process and growth.
• 超驗的觀點：研究者提出問題、研究目的在移除錯誤信念、改變並超越目前社會情境
• Qualitative social work researchers usually follow a non-linear path, going wherever
the setting takes them, emphasizing more imprecise measures, and using intuition or
insight.
• 非線性路徑：擴散的循環取向，形成整體感受、建構意義為主，借用文學的一些手法

Triangulation 三角檢核法
• Triangulation involves multiple measures of the same phenomenon. 測量工具
• Triangulation of theory is multiple theoretical perspectives used during planning or
analysis of the setting. 理論
• Triangulation of methods is the mixing of multiple styles that may be qualitative and
quantitative. 方法
• 觀察者

Objectivity and Integrity 客觀與誠實（量化）


• Quantitative social work researchers use more mechanical, standardized procedures to
gather ‘objective’ data from a distance, and they measure using numbers and analyze
using mathematical techniques.
• 強調客觀、採用機械化技巧（複製、標準化、以數字測量、統計分析）、排除人為因素

客觀與誠實（質化）
• Qualitative social work researchers emphasize the human factor and firsthand
‘subjective’ knowledge. They immerse themselves in the setting, rather than
remaining distant, they emphasize trust.
• 強調可信賴性 (trustworthiness) 、對自我的介入保持敏感、確保研究活動是可信的 (dependable)、及可靠的
(credible)

二、量化設計議題

量化設計課題 (1)變項與假設的語言
• Variables 變項 are measures of a concept that take two or more values, such as male or
female.
• Attributes 屬性 are the values or categories of a variable. Male is an attribute of the
variable gender.
• The independent variable 自變項 is the presumed cause, and the factor that precedes the
effect in time.
• The dependent variable 應變項 is the effect or outcome that is presumed to be impacted,
and follows the cause in time.
• Intervening variables 中介變項 come between the independent and dependent variables.
量化設計課題 (2) Causal Hypotheses 因果假設
• The causal hypothesis has at least two variables.
• It expresses a cause-effect relationship.
• It is expressed as a prediction or expected future outcome.
• It is logically linked to theories and the research question.
• It is falsifiable or capable of being tested against empirical evidence.

假設的檢定與精進
• 知識的演進很少植基於單一假設的檢定，易使研究歷程產生扭曲的圖像
• 一個假設必須經過多次檢定後才會受到廣泛的接受－可複製性原則
• 假設檢定隨時間運作的過程 圖 6-2

假設的類型
• 虛無假設：基於否證論，藉推翻虛無假設使對立（研究）假設得以保留在論點中
• 雙載假設 (double-barreled hypothesis) ：在一個假設中放進兩個關係
• 交互作用效果：結合兩個變項預測應變項
• 如：貧窮與青少年高度聚集會提高財物犯罪率（可能是雙載或交互）

Null Hypothesis 虛無假設


• The null hypothesis normally states the negation of the relationship proposed in the
research hypothesis.
• If a research hypothesis stated: “Males are more likely than females to repeat juvenile
offenses,” the null would state: “There is no relationship between the sex of the
offender and number of repeat offenses.”

解釋的切入點 Aspects of interpretation


• 分析層次：理論解釋所指涉的社會實在層次（如鉅觀或微觀）
• 分析單位：研究者在觀察（測量）時所使用的單位類型（個人、團體、組織、社會類別）
• 不同理論或觀察技術著重的分析單位不同
• 微觀分析層次適合以個人為分析單位
• 鉅觀分析層次適合以社會類別與機構為分析單位
• 調查與實驗法適合以個人為分析單位

Logical Errors in Causal Explanation 因果解釋中的邏輯錯誤


套套邏輯
• Tautology is a form of fallacious circular reasoning. Sally is conservative because she
believes in less regulation. While this may be true, her conservativeness is not caused
by her attitude toward regulation.
• 循環推理，無法接受實徵資料的考驗
• 陳述中並無解釋，只是一種再敘述

目的論
• Teleology are causes that are associated with an ultimate purpose or goal. “Crime
occurs because humans are destined to be evil …” is teleological. This violates the
principle that the cause precedes the effect in time.
• 事情由終極的目標或目的所導引（將目標當作動機）
• 欠缺一個真正的自變項
• 違反時間順序的因果條件（原因含糊不清）
• 無法以實徵方式進行測量 （humans are evil）

區位謬誤
• Ecological fallacy arises from a mismatch of units of analysis when moving from a
larger to a smaller unit. Knowing the vote totals for a particular polling district does
not allow one to conclude how any one person voted.
• 將較高分析單位（集合體）收集的資料用來對較低單位（個人）作結論

化約論
• Reductionism is a mismatch in units of analysis when moving from smaller to larger.
Knowing how some individuals in an organization behave does not allow one to
conclude the entire organization behaves in a certain way.
• 以個體微觀層次的理由或行為去解釋如社會制度等鉅觀層次的行動
• 未等化謬誤

虛假關係
• Spuriousness occurs when two variables are associated but not causally related. The
statistical association between air conditioning and ice cream cones sales is spurious,
not causal.
• 兩個變項的關聯並非因果關係，而是由看不見的第三因素所造成
• 例如：身高與足球喜好的關聯，第三因素可能是性別

三、Ways to Select a Topic 選擇主題的方式：量化研究


• Personal experience
• Curiosity based on current events
• State of knowledge in field
• Solving a problem
• Social premiums
• Personal values
• Everyday experiences

Narrowing Down a Topic 縮減題目的範圍


• Examine the literature.
• Talk with others.
• Limit analysis to a specific context.
• Define the aim and outcomes.

Research Questions 研究問題


• Bad: Is child abuse a problem?
• Good: Has the incidence of child abuse changed in the last three years in Missouri?
• Bad: Do social workers affect delinquency?
• Good: Has the incidence of delinquency changed among caseloads of MSW level
social workers compared to BSW workers in Milwaukee?

Steps in Research 量化研究步驟


• Choose a topic.
• Refine topic into research question.
• Design study.
• Collect data.
• Analyze data.
• Interpret data.
• Inform others.

四、質化研究設計議題
個案與脈絡的語言 The language of cases and contexts
• 由研究對象身上去借得想法，或是在特定脈絡或自然環境中，檢視一個特定個案時發展新的構想
• 不急於將對社會世界的想法或觀點轉化成一般變項以形成假設

紮根理論
• Grounded Theory occurs when a qualitative researcher begins with a question and
little else. Theory is developed inductively as experiences unfold and data are
collected. The context or setting is critical.
• 使用一套有系統的程序，去發展出對於某個現象歸納而得的理論

脈絡是關鍵 The context is critical


• 社會行動或陳述的意義必須視其出現的脈絡而定
• 研究者注意「之前發生什麼」「研究焦點的周遭有什麼」等

雙手萬能
• Bricolage occurs when a qualitative researcher draws on a variety of sources, making
do with whatever is at hand, mixing diverse approaches, and assembling bits and
pieces.
• 就地取材、窮則變通的能力，無法預先計畫

重視個案與過程 Focus on the case and process


• 研究者關注事件的先後順序
• 長期檢視一個或一組個案，觀察一個議題的醞釀成形、衝突或併發、或是一個社會關係的發展
• 探測過程及過程中的因果關係
詮釋
• Interpretation is the act of assigning significance or coherent meaning by translating
and making understandable the data that have been collected.
• 對於所獲資料賦予重要的或貼切的意義
• 第一級詮釋：瞭解研究對象的理由、動機與意義
• 第二級詮釋：對第一級詮釋有所發現或重新建構
• 第三級詮釋：與一般性理論做連結

五、雕琢研究主題 Refining research topics：質化研究


• 研究問題的聚焦多在展開資料收集後進行
• 典型的研究問題如：
某個狀態或社會情境是如何萌芽的？
這個狀態或情境在當時是如何維持的？
其改變、發展或運作力為何？

研究範例
量化 質化
實驗 田野研究
調查研究 歷史比較研究
內容分析 菁英研究

You might also like