Last minute Qs Flashcards
What two stages of interpretation are important to hermeneutics?
1.Uncovering interpretation (finding how others have categorised the world)
2.Assigning interpretation (creating new categories)
Where is induction used?
Qualitative research - moving from the domain of observations, to making findings, to creating theories
Where is deduction used?
Quantitative research - moving from a theory, to creating a hypothesis and testing it, to validation
What is most important to action research?
Solving problems! Studying real life problems without coming up with solutions is unhelpful
What is an artifact in IT-research?
Something created by human beings that cannot exist without human involvement - anything from a computer to a set of values
What kind of artefacts does design research create?
Constructs, methods, models, instantiations, theories
In technical action research, what are the three roles of the researcher?
Designer, helper, researcher
What is the difference between action research and technical action research?
In action research the researcher wants to solve a problem for a client, in technical action research the problem is only a way to learn about a specific technique
Four important factors to ethics in research?
Reliability, honesty, respect, accountability
What two types of interpretation are important to IS (Information Systems) studies?
Both positivist and interpretive
Why is deduction an important tool to computer science?
The scientific foundation is based on logic and discrete mathematics
What is the most common research approach in computer science?
Controlled experiments with statistical analysis and inference
What is IT?
The study of systemic approaches to develop secure computing technologies that fulfill our goals
Are literature mapping studies quantitative or qualitative in nature?
Qualitative - they are about making sense of the material and creating a structure
Are systematic literature reviews qualitative or quantitative in nature?
Quantitative to a large extent, validity threats need to be resolved
What are the three high-level phases of a systematic literature review?
Planning, conducting and reporting the review
What is the difference between a primary and a secondary study?
A secondary study is the literature review itself; primary studies are the studies used as sources within the literature review
What is the primary benefit of using surveys as a research method?
Asking many people in a time-efficient way, with flexible delivery options
What is the downside of using surveys as a research method?
No chance to ask follow-up questions or verify how questions are interpreted
What are the three important qualities needed in a survey?
-Relevance to research Qs
-Easy to interpet
-Interpreted the same by all participants
What are the three steps of survey development?
-Define variables
-Develop Qs to measure variables
-Pilot and validate
What are different types of questions that can be used in a survey?
Knowledge, Likert scale, open/closed
What are the two ways we can validate our survey questions/approach?
-Expert review
-Pilot study
In the context of validating surveys, what is a cognitive interview?
A hybrid between an expert review and a pilot study where participants do the survey while monitored by a researcher
What is necessary to do before a survey to uphold ethics requirements?
For participants to sign a consent form outlining the purpose and methodology of the survey, along with information about who’s behind the survey
What is a sample?
A subset of our population we can assume is representative of the population at large
How is stratified sampling different from random sampling?
The population is narrowed down by criteria like age/gender and a random sample is taken from that sub-population
What is the difference between convenience sampling and purposive sampling?
Asking whoever is close at hand vs asking the most high-value individuals for the area in question
Three types of bias in surveys?
-Confirmation bias
-Survey fatigue bias
-Question bias
In the context of experiment design, what are design points?
Unique combinations of independent variable settings
An independent variable needs to have two what?
Levels; these are the test conditions
What is a control variable?
A variable that is kept constant while testing the effect of the independent variable
What are the two ways of assigning conditions to participants?
Within-subject testing where each participant is tested on each condition, or between-subject testing where each participant is tested on one condition
What is the problem of using too many participants in an experiments?
Achieving statistical significance for effects of no practical value
What are usability lab-studies?
Participants are brought into a lab, one-on-one with the researcher and given scenarios that lead to tasks relevant to a product/service
What are internal validity threats?
Reasons why inferences about the causal relationship between two variables might be incorrect
What are external validity threats?
Reasons why inferences about why study results would hold over variations in inviduals/setting are incorrect
What is a good thing to do after the experiment to get a better estimate of experimental error?
Replication
Academic writing should be…
Clear, concise, objective
How do we test if our results are statistically significant?
By assessing if the probability of observing the results we have achieved, even if our sample is not representative of the population at large, is above our selected significance level (95%)
What is the p-value?
The probability of getting these results even if the null hypothesis is true
What is α?
Significance level
If p-value is less than α, what do we do?
We reject the null hypothesis
What are the requirements for a t-test?
-Interval or ratio data
-Small samples require normal distribution, medium samples close to normal distribution but large samples work regardless of distribution
In a distribution test, what does it mean for a p-value to be above 0.05?
Normal distribution
What is a Rank-Sum test?
-aka Mann-Whitney U-test
-Used to test independent variables
-Ranks individual MEASURES instead of using means
What are the requirements for a Rank-Sum test?
-At least ORDINAL data
-Independent variable divided into two groups
-Distribution should ideally be similar
In what situation would we use a Sign Test?
Cases with paired observations where t-test is not okay
What are the requirements for a CHI2 test?
-Works for all data levels
-No cells can have an expected count of 0
-At least 80% of the cells must have a count of 5 or more
What does it mean to find a linear correlation between two variables?
For one variable to increase with the other
Six steps of thematic analysis?
1.Familiarising
2.Generating codes
3.Searching for themes
4.Reviewing themes
5.Defining and naming themes
6.Producing report