W2: Ethics, Survey Design and Deriving Scales Flashcards
What are the three core principles of human ethics?
Respect/Autonomy, Beneficence, Justice
What does Respect/Autonomy entail in ethical research?
Participants must be able to make autonomous, reasoned decisions about their participation
What is Informed Consent?
Participants are informed about the purpose of the study, risks/benefits, rights, and debriefing
What is the purpose of debriefing in research?
To inform participants about the study after their participation
What special considerations are there for special populations in research?
Ensuring that participation is autonomous and that risks/rewards are considered
What does Beneficence require in research?
Utilize benefits and minimize risks of the research
What should research address according to the principle of Beneficence?
Research that is relevant to a societal need and advances scientific theory
What is required to ensure participant beneficence?
Protect participants from psychological and physical harm and maintain privacy and confidentiality
What is the principle of Justice in ethical research?
Selection of participants and risks/benefits must be equitable
What should inclusion/exclusion criteria be based on?
Research questions or limitations such as age, gender, socioeconomic status, ethnicity, and sexual orientation
True or False: Ethical decisions must avoid all costs and injustices.
False
What is the significance of the Treaty of Waitangi in ethical research at VUW?
It must be considered alongside the principles of respect/autonomy, beneficence, and justice
What are core application requirements for obtaining ethics approval?
Include copies of all materials, description of confidentiality, benefits/risks, compensation, and data storage
What is the difference between morality and ethics in research?
Morality is personal values, while ethics are standards set by committees and associations
Fill in the blank: Ethical issues in research include _______, Beneficence, and Justice.
Respect/Autonomy
What does Research Beneficence refer to in science in the media?
It applies to ensuring ethical standards in research reported in the media
What is the goal of applying participant and survey characteristics?
To create ‘good’ questions
What does measurement of item reliability involve?
Assessing the consistency of items within a scale
What are common biases in surveys that affect validity and reliability?
Response Biases, Demand Characteristics, Social Desirability
What are scale anchors used for?
To provide strong question phrasing and avoid ceiling and floor effects
What are examples of scale anchors?
‘not at all’/‘always’, ‘strongly disagree’/‘strongly agree’
What is Acquiescence Bias?
The tendency for people to agree more than they disagree
In which cultures is Acquiescence Bias stronger?
Collectivistic cultures
What can help mitigate Acquiescence Bias?
Using multiple items to average into one scale
What does Cronbach’s Alpha test measure?
Item reliability within a scale
What is considered an acceptable Cronbach’s Alpha score?
Above .70
What is Priming in the context of surveys?
Exposure to a question influences subsequent answers
What is an example of Affective Priming?
Answering questions about values, morals, or attitudes
What should researchers do to deal with Priming?
Move impactful questions to the end of the survey or use distraction tasks
What is a common mistake in survey question design?
Double-Negatives or Double-Barreled Questions
What can leading questions do to survey measures?
Make measures totally invalid
What is the effect of Demand Characteristics and Social Desirability on response bias?
Increase bias but lower validity
What should good survey questions adjust for?
Psychological biases in the participant and in the survey
Fill in the blank: Acquiescence Bias is the tendency to agree more than to _______.
disagree
What is a method to counter Acquiescence Bias?
Remove the mid-point of the scale
What can high reliability and low validity indicate in survey data?
The measures are consistent but not accurately capturing the intended construct
In an online survey, what does a countdown timer help adjust for?
Social desirability/Demand characteristic biases
What is the main purpose of Principal Components Analysis (PCA)?
To identify which questions group together and which provide unique information.
What is the first step in developing a scale?
Operationalization.
What should you ensure when designing questions for PCA?
Balance question wording and valence.
What is the recommended sample size for PCA?
Greater than 200 and more than 5 people per item.
What does PCA look at to summarize measures?
Patterns of data and fitting a series of lines called ‘Principal components’.
What rotation method should you choose in PCA when components are likely to be correlated?
‘Oblimin’ rotation.
What is the significance of the Bartlett test in PCA?
It assesses sphericity and needs to be significant.
What does the Kaiser-Meyer-Olkin (KMO) measure assess?
Sampling adequacy, needing to be above 0.5.
What indicates that it is appropriate to conduct a PCA?
Significant Bartlett’s test and a KMO measure above 0.5.
What does the scree plot help determine?
The number of components to retain.
In the context of PCA, what does ‘cross-load’ mean?
An item loads on multiple components.
What should be done if items cross-load in PCA?
Consider re-starting or removing the item from the scale.
What is one key indicator to look for in the scree plot?
The point where the plot drops to the flat part (the ‘plateau’).
What is meant by ‘construct validity’ in PCA?
The grouping of items reflects the theoretical constructs being measured.
What should you do with items that have negative loadings when forming a scale?
Reverse-code those items.
What should be included in the report after conducting PCA?
Descriptive statistics and reliability for the new scale.
Fill in the blank: The two components summarizing reasons for crying were labeled as _______ and _______.
self-focused reasons, other-focused reasons.
True or False: An item loading on three different components should always be kept in the scale.
False.
What is the goal of generating as many questions/items as possible in PCA?
To select items appropriately for your aim and sample.
What does a high KMO score indicate?
Very good sampling adequacy.
What is the purpose of averaging items in each component?
To create a reliable scale.
What should be reported in an APA table after PCA?
Item labels and loadings.
What does it mean if components are grouped by something other than theorized?
It may indicate a lack of construct validity.
