研究设计与效度问题

参考文献：

1. Reis H T, Judd C M. Handbook of research methods in social and personality psychology[M]. Cambridge University Press, 2000.

• Bryman A. Social research methods[M]. Oxford university press, 2015.

1.研究目的有三种：描述性研究；因果推论；理论解释。

• 描述性研究：验证一种现象或关系的存在，描述世界的某种状态，包括在某一时间与空间范围内特殊事件的发生频率以及评估特色状态或情境之间的关联程度。
  现场研究OR实验室
• 因果推论：验证变量之间的因果联系，如果X那么Y。找出那些可用于控制或者操作的动因，以便引起结果变量的变化。

任何引起世界某一种状态发生变化的干预，都将会随之带来研究结果状态的变化。

验证，not描述

自变量被操纵——实验或准实验
时间滞后，假设自变量的变化先于结果变量——自然情况

因果推测的效度很大程度上取决于研究者对自变量的定义！

• 理论解释：用于基础理论检验，仅仅了解因果的存在是不够的，要进行解释，验证X与Y之间存在的干预中介过程。才能反映因果关系的“本质概念”。

2.效度：Campbell 1957，1963，1966区分除了经典的内部效度（internal validity）和外部效度（external validity）。Cook and Campbell 1979将外部效度进一步区分，引入构念效度。

• 内部效度：因变量的改变是不是由自变量的改变引起的
  A concern with the question of whether a finding that incorporates a causal relationship between two or more variables is sound.

• 外部效度：在不同的被试群体中、在不同的实验条件下、使用不同的方法，是否可以获得相同的因果关系
  A concern with the question of whether the results of a study can be generalized beyond the specific research context in which it was conducted.
  （1）稳健性：实验可以重复吗？可以推论至多个人群和情境；可推论至某个特定人群。

有关外部效度或者推论概括性的各种问题实际上就是自变量（实验处理）和情境变量（被试选择、被试的经历以及研究设置等）之间相关关系的问题。对外部效度的本质关注是寻找那些限制和影响研究中因果关系的调节因素，一旦能够辨识出潜在的调节因素，就可以通过系统地改变这些因素，看他们是否导致结果变量的变化，从而验证研究效果的稳健性。Cook， Campbell 1979

（2）生态效度 Ecological validity : 它有代表性吗？研究结果变量的改变是否如同它在现实中发生的改变一样？
A concern with the question of whether social scientific findings are applicable to people's everyday, natural social setting.
引导出因果关系的实验情境与现实生活中包含同样因果关系的情境不一定要表面上看起来十分类似，才能证明该因果是有效的。
（3）相关性：它重要吗？研究结果对于问题解决或者提高生活质量是否具有潜在的帮助作用和适用性？
是一种最终形式的推论概括性，其差异表现在程度而非类型上。
基础研究结果及其应用之间的联系通常是间接的和累进的，而不是即刻就能实现的。

• 构念效度：在特定研究中使用的特殊方法和操纵条件下得到的因果关系，可以被推论到研究试图表征的理论构念及过程的程度。
  The correspondence between a construct and the operational procedure to measure or manipulate that construct. Two indicators, convergent validity（Cronbach's alpha) and discriminant validity(Factor analysis).
  Measure validity : The degree to which a measure of concept truly reflects that concept. See also face validity and concurrent validity.
  Face validity :A concern with the question of whether an indicator appears to reflect the content of the concept in question.
  Concurrent validity : one of the main approaches to establishing measuring validity. It contains relating a measure to a criterion on which cases are known to differ and that is relevant to the concept in question.

Concurrent validity measures how well a new test compares to a well-established test. It can also refer to the practice of concurrently testing two groups at the same time, or asking two different groups of people to take the same test.
Example 2: Concurrent validity can also occur between two different groups. For example, let’s say a group of nursing students take two final exams to assess their knowledge. One exam is a practical test and the second exam is a paper test. If the students who score well on the practical test also score well on the paper test, then concurrent validity has occurred. If, on the other hand, students who score well on the practical test score poorly on the paper test (and vice versa), then you have a problem with concurrent validity. In this particular example, you would question the ability of either test to assess knowledge.
[statistic show to](http://www.statisticshowto.com/concurrent-validity/）

Predictive validity: tells you how well a certain measure can predict future behavior.

3.效度的针对性

• 描述性研究：生态效度最为重要，描述的是否为真实的世界？
• 应用研究：稳健性特别重要，可否复制和推广
• 设计解决应用问题的方法：稳健性，生态有效性，结果是否脆弱，推论范围是否狭窄
• 理论解释：构念效度比外部效度更重要