PSY2013 Lecture 06 Notes
PSY2013 Research Methods in Psychology
Ch09 Non‐Experimental Design I: Survey Methods
Ch10 Non‐Experimental Design II: Observational and Archival Methods
Ch11 Quasi‐Experimental Designs and Applied Research
Ch12 Small N Designs
Non-Experimental Design 非实验性设计
Learning Objectives 学习目标
Explain why sampling issues are more relevant for survey research than for most other research in psychology
解释为什么抽样问题对调查研究比对心理学中的其他研究更相关Articulate the principles of good survey construction
阐述构建良好调查的原则Explain the problems that can make it difficult to interpret survey data
解释导致难以解释调查数据的问题Describe four ways to collect survey data and list the advantages and disadvantages of each
描述四种收集调查数据的方法，并列出每种方法的优缺点Describe three varieties of probability sampling and know when each is used
描述三种概率抽样，并知道每种抽样的使用情况Distinguish between positive and negative bivariate correlations, create scatterplots to illustrate them
区分正相关和负相关双变量，创建散点图来说明它们Calculate a coefficient of determination (r²) and interpret its meaning
计算决定系数 (r²) 并解释其含义Understand how a regression analysis accomplishes the goal of prediction, and distinguish between simple linear regression and multiple regression techniques
理解回归分析如何实现预测目标，并区分简单线性回归和多元回归技术Understand how directionality can make it difficult to interpret correlations and how a cross-lagged panel study can help with the directionality problem
理解方向性如何使解释相关性变得困难，以及交叉滞后面板研究如何帮助解决方向性问题Understand the third variable problem and how such variables can be evaluated and controlled through a partial correlation procedure
理解第三变量问题，以及如何通过部分相关程序评估和控制这些变量Distinguish between mediators and moderators within the context of understanding third variables in a correlation
在理解相关性的第三变量时区分中介变量和调节变量
Survey Research 调查研究
Origins of Survey Research 调查研究的起源
- Early Surveys 早期调查
- Darwin’s study on facial expressions of emotion 达尔文对情感面部表情的研究
- Galton’s work on the origins of scientific interests 高尔顿对科学兴趣起源的研究
- Hall’s research on the contents of children’s minds 霍尔对儿童思维内容的研究
Sampling Issues 抽样问题
- Biased vs. Representative Samples 有偏见的样本 vs. 代表性样本
- Non-Probability Sampling 非概率抽样: Convenience sampling, quota sampling, etc.
- Probability Sampling 概率抽样: Simple random sampling, stratified sampling, cluster sampling
- Self-Selection Bias 自我选择偏差: Example of 1936 Literary Digest election poll 1936年《文学文摘》的选举民调例子
Creating an Effective Survey 创建有效的调查
Types of Survey Questions 问题类型
- Open-Ended Questions 开放式问题: Allows respondents to answer in their own words
- Closed Questions 封闭式问题: Provides predefined response options
- Likert Scales 李克特量表: Measures attitudes by asking respondents to rate their agreement on a scale
Survey Wording 问卷措辞
- Avoid ambiguity 避免模糊
- Avoid double-barreled questions 避免双重问题
- Avoid biased and leading questions 避免有偏见和引导性的问题
Collecting Survey Data 收集调查数据
- Methods 方法
In-Person Interview Surveys 面对面访谈
- Advantages: Comprehensive data, follow-ups possible
- Disadvantages: High cost, interviewer bias
Mailed Written Surveys 邮寄书面调查
- Advantages: Ease of scoring
- Disadvantages: Low response rate, nonresponse bias
Phone Surveys 电话调查
- Advantages: Cost-effective, efficient
- Disadvantages: Must be brief, lower response rate
Electronic Surveys 电子调查
- Advantages: Cost-effective, efficient
- Disadvantages: Sampling issues, ethical concerns
Ethical Considerations 伦理考虑
- Issues 问题
- Informed consent 知情同意
- Privacy and confidentiality 隐私和保密
- Use and misuse of survey data 调查数据的使用和滥用
Analyzing Data from Non-Experimental Designs 分析非实验性设计的数据
Correlation: Describing Relationships 相关性：描述关系
Types of Correlations 相关类型
- Positive Correlation 正相关: Higher scores on one variable associated with higher scores on another
- Negative Correlation 负相关: Higher scores on one variable associated with lower scores on another
- No Correlation 无相关性: No consistent relationship between two variables
Scatterplots 散点图
- Visual representation of data from two variables
- Helps in identifying the type and strength of correlation
Correlation Coefficients 相关系数
Statistical Tests 统计检验
- Pearson’s r: Measures linear correlation between two variables
- Spearman’s rho: Non-parametric measure of rank correlation
- Phi Coefficient: Used for binary variables
Coefficient of Determination 决定系数 (r²)
- Proportion of variability in one variable that can be explained by variability in another variable
- Example: r = .60 → r² = .36 (36% of variability can be explained by the relationship)
Regression: Making Predictions 回归：做出预测
Simple Linear Regression 简单线性回归
- Predicting the value of one variable based on the value of another
- Equation: Y = a + bX
- Y: Criterion variable (dependent variable)
- X: Predictor variable (independent variable)
- a: Y-intercept
- b: Slope of the line
Multiple Regression 多元回归
- Predicting the value of a criterion variable based on multiple predictor variables
- Allows for the weighting of each predictor variable’s influence
Interpreting Correlational Results 解释相关结果
Directionality Problem 方向性问题
- Correlation does not imply causation
- Cross-lagged panel studies can help determine the directionality
Third Variable Problem 第三变量问题
- An uncontrolled third variable could influence both correlated variables
- Partial correlation can help control for third variables
Mediators and Moderators 中介变量和调节变量
- Mediator 中介变量: Explains how or why a relationship exists between two variables
- Moderator 调节变量: Explains under what conditions the relationship exists
Observational Research 观察研究
Varieties of Observational Research 观察研究的种类
Naturalistic Observation 自然观察
- Describing behaviors in their natural settings without intervention
- Examples: Snack selection at movie theaters, gender differences in fighting behaviors at a bar
Participant Observation 参与观察
- The researcher joins the group being observed
- Example: Festinger’s study of a cult
Challenges in Observational Research 观察研究的挑战
- Absence of Control 缺乏控制
- Can falsify strong claims but cannot establish causality
- Observer Bias 观察者偏见
- Use behavior checklists and ensure interobserver reliability
- Participant Reactivity 参与者反应
- Use unobtrusive measures to minimize reactivity
- Ethics 伦理
- Issues with consent and privacy
Analyzing Qualitative Data from Non-Experimental Designs 分析非实验性设计中的定性数据
- Thematic Analysis 主题分析
- Identifying patterns in qualitative data
- Steps:
- Get familiar with the data 熟悉数据
- Code 编码
- Search for themes 搜索主题
- Review themes 审查主题
- Define and name themes 定义和命名主题
- Write report 撰写报告
Archival Research 档案研究
- Using Previously Collected Data 使用以前收集的数据
- Often involves content analysis
- Susceptible to missing data and bias
- Analyzing Archival Data 分析档案数据
- Factor analysis identifies clusters of related items
- Meta-analysis evaluates effect sizes across multiple studies
Quasi-Experimental Designs 准实验性设计
Learning Objectives 学习目标
Identify the dual functions of applied research
识别应用研究的双重功能Understand why applied research has always been an important element in American psychology
理解为什么应用研究一直是美国心理学中的重要元素Define translational research and explain how psychological research can translate into applied settings
定义转化研究并解释心理学研究如何转化为应用环境Identify the design and ethical problems associated with applied research, especially if that research occurs outside of the laboratory
识别与应用研究相关的设计和伦理问题，特别是当该研究在实验室外进行时Identify the defining feature of quasi-experimental designs, and recognize which designs appearing in earlier chapters were quasi-experimental
识别准实验性设计的定义特征，并认识到前几章中出现的哪些设计是准实验性的Describe the features of a nonequivalent control group design and understand why this design is necessarily confounded
描述不等价对照组设计的特点，并理解为什么这种设计必然混淆Understand why matching nonequivalent groups on pretest scores can introduce a regression artifact
理解为什么在前测分数上匹配不等价组可以引入回归伪影Describe the features of interrupted time series designs and understand how they can be used to evaluate trends
描述中断时间序列设计的特点，并理解它们如何用于评估趋势Describe variations on the basic time series design
描述基本时间序列设计的变体Describe the strategies that can be used complete a needs analysis in a program evaluation project
描述在项目评估项目中完成需求分析的方法Understand the purposes and the procedures involved in various program evaluation procedures
理解各种项目评估程序的目的和程序Identify and describe the ethical problems that often accompany program evaluation research
识别并描述经常伴随项目评估研究的伦理问题
Beyond the Laboratory 超越实验室
Applied Research 应用研究
- Solves real-world problems and increases basic knowledge 解决现实世界的问题并增加基础知识
- Example: Nutrition labels’ effect on health decisions 营养标签对健康决策的影响
Design Problems in Applied Research 应用研究中的设计问题
- Ethical dilemmas, trade-off between internal and external validity 伦理困境，内部和外部效度之间的权衡
- Problems with creating equivalent groups, uncontrolled order effects 创建等价组的问题，未控制的顺序效应
Quasi-Experimental Designs 准实验性设计
Features 特点
- No causal conclusions, less than complete control, no random assignment 无因果结论，控制不完全，没有随机分配
- Examples from prior chapters: single-factor nonequivalent groups designs, ex post facto factorial designs, P x E factorial designs 以前章节中的例子：单因素不等组设计，事后因素设计，P x E 因素设计
Nonequivalent Control Group Designs 不等价对照组设计
- Include pretests and posttests 包括前测和后测
- Random assignment not possible for practical reasons 由于实际原因，无法进行随机分配
Interrupted Time Series Designs 中断时间序列设计
Features 特点
- Evaluates overall trends 评估总体趋势
- Example: Effect of incentive plan on productivity 激励计划对生产力的影响
Variations 变体
- Adding a control group 添加对照组
- Adding a second treatment at a different time 在不同时间添加第二个处理
- Adding a second DV that is not expected to be influenced by the program 添加一个不受程序影响的第二个因变量
Program Evaluation 项目评估
Steps 步骤
Needs Analysis 需求分析
- Census data, surveys of resources and potential users, focus groups 人口普查数据，资源和潜在用户调查，焦点小组
Formative Evaluation 形成性评估
- Evaluating the program while it is in progress 在项目进行时进行评估
- Program audit, pilot study 项目审计，试点研究
Summative Evaluation 总结性评估
- Evaluating program effectiveness 评估项目效果
- More threatening than formative evaluation 比形成性评估更具威胁性
- Use of quasi-experimental designs 使用准实验性设计
Cost-Effectiveness Analysis 成本效益分析
- Comparing the costs and effectiveness of different programs 比较不同项目的成本和效果
Ethical Issues 伦理问题
- Concerns 关注点
- Consent, confidentiality, perceived injustice 知情同意，保密性，感知的不公正
- Stakeholder conflicts 利益相关者冲突
Small N Designs 小样本设计
Learning Objectives 学习目标
Describe examples of classic studies in psychology’s history, all using single individuals or a small number of participants, with additional subjects used for replication
描述心理学历史上经典研究的例子，这些研究都使用单个个体或少数参与者，并使用额外的受试者进行复制Explain how grouping data from large numbers of participants can yield misleading conclusions about behavior
解释将大量参与者的数据分组如何得出关于行为的误导性结论Describe some practical reasons for doing small N research
描述进行小样本研究的一些实际原因Describe B. F. Skinner’s basic philosophy about how best to conduct research—the experimental analysis of behavior
描述B. F. 斯金纳关于如何最好地进行研究的基本哲学——行为的实验分析Describe the essential components of any single-subject design
描述任何单个受试者设计的基本组成部分
The Experimental Analysis of Behavior 行为的实验分析
- Operant Conditioning 操作性条件反射
- Behavior conditioned by consequences (e.g., reinforcement) 行为由结果（如强化）条件化
- Primary DV: Rate of response 主要因变量：反应率
Small N Designs in Applied Behavior Analysis 应用行为分析中的小样本设计
Single-Subject Designs 单受试者设计
- Importance of operational definitions 操作定义的重要性
- Baselines established 建立基线
- Introducing treatment 引入处理
Withdrawal Designs 撤回设计
- A-B Design: Baseline (A), Treatment (B)
- A-B-A Design: Baseline (A), Treatment (B), Withdrawal (A)
- A-B-A-B Design: Baseline (A), Treatment (B), Withdrawal (A), Treatment (B) again
Multiple Baseline Designs 多基线设计
- Variations include multiple subjects, multiple behaviors, multiple environments 变体包括多个受试者，多个行为，多个环境
Changing Criterion Designs 变化标准设计
- For behaviors that must be changed gradually 对必须逐步改变的行为
Alternating Treatments Designs 交替处理设计
- Compares the relative effectiveness of two or more treatments 比较两种或多种治疗的相对效果
Case Study Designs 案例研究设计
Detailed Analysis 详细分析
- Incorporates other methods (e.g., observation, psychometric tests) 融合其他方法（如观察、心理测量测试）
Evaluating Case Studies 评估案例研究
- Strengths: Level of detail not found elsewhere 优势：独有的详细程度
- Limitations: Limited control, external validity issues 限制：有限的控制，外部效度问题
Summary 总结
Surveys are used to obtain information primarily about attitudes, beliefs, opinions and/or projected behaviors.
调查主要用于获取关于态度、信仰、观点和/或预测行为的信息。Correlation and regression are statistical techniques that can be used to analyze data from non-experimental designs, including survey data.
相关和回归是统计技术，可以用来分析非实验性设计的数据，包括调查数据。Observational methods are used to assess natural behaviors either in their natural setting or in a lab setting.
观察方法用于评估自然环境或实验室环境中的自然行为。Quasi-experimental designs are used when participants cannot be randomly assigned to groups, and include: nonequivalent control group and interrupted time series designs.
当参与者不能随机分配到组时使用准实验性设计，包括：不等价对照组设计和中断时间序列设计。Program evaluation is used to assess the effectiveness of large-scale programs.
项目评估用于评估大规模项目的效果。Small N designs are used when individual-subject validity is weak and when studying individuals from rare populations.
当个体受试者效度较弱时，以及研究稀有人群时使用小样本设计。Case studies are in-depth analyses of one or a few people, usually to develop treatment plans or obtain knowledge.
案例研究是对一个或几个人的深入分析，通常是为了制定治疗计划或获取知识。
PSY2013 Lecture 06 Notes