forskningsmetoder

implicit knowledge informal, in-head

explicit knowledge formal, documents, databases

descriptive/categorial knowledge most fundamental, describe phenomen with its characteristics

explanation/declarative knowledge why a phenomen is in a certain way, cause-effect-relation explained

evaluation/value knowledge knowledge about desired - given specific circumstances, goals

normative/prescriptive knowledge guiding knowledge, instructions

positivism how reality is! highly structured, quantifiable observations that lend to statistical analysis, complexity reduced to generalization

realism object exist independently of our will & knowledge

pragmatism focus on the practical consequenses of actions, if concept works in practise - it is considered true

Interpretation & hermeneutics to understand phenomens from another person's perspective

Induction start observe, then theory

deduction start with theory, then validation/proof

action research consult, support practise, tech in social context

design science research response to an it problem, artifact, new solutions

technical action research tech-driven, solve a lot of problems, designer, helper, researcher

what is good research orginality, reliability, communication, relevance, rigor

why literature review need to know what has been solved in order to motivate research gap

literature process define topic, limit search, test search, review results, conduct search, manage results, evaluate results, report

systematic literature review focused on gathering evidence and resolve validity threats

bad literature review lack of focus, organization, depth, bad sources, exist research

survey purpose collect data from a large sample of respondents

survey steps define variables to measure, develop questions, validate

piloting distribute survey to a small set of respondents

stratified sampling practical random approach, sub-groups based on criterias such as age, gender

confirmation bias agree rather than disagree

when should we use experiments need to identify dependable relationship between cause & effect

casual relationship when the cause precedes/relates to the effect

independent variable cause, variable manipulated under experiment

dependent variable effect, depends on the independent variable

hypothesis precise problem statement that could be tested

confounding variable not controlled but can affect outcome

control variable not under investigation, kept constant, font size

within-subjects each p tested on each condition

between-subjects each p tested on one condition only

replication repetition of entire experiment

purpose statistic analyze if our results is accurate

confidence interval maesure of how confident we are in our results

Inferential statistics how we draw conclusions from data - correlation, hypothesis

data level determines what type of measurements & procedures we can apply

nominal data categorial data with no meaningful distance, animal

nominal data descriptives frequencies & percentages, mode

ordinal data ordered data with undefined length between steps, tasty

ordinal descriptives frequencies & percentages, median

interval & ratio equal distance between steps, parametric

interval & ratio descriptives mean, standard deviation

mean medelvärde

standard deviation hur spridda värdena är runt medelvärdet

hypothesis testing goal to determine if we can reject null hypothesis in favor of the one we defined

statistically significance - hypothesis testing acheived when a result is very unlikely given the null hypothesis is true

p-value p under a = s results - reject null

t-test one sample if mean of one sample is different from reference value

t-test independent sample test if the mean of one sample is different from mean of another independent sample

t-test paired sample test if the mean of 2 measures from the same sample differs

normal distribution test curves, 0,05

rank-sum test independent variables, ranking indicidual measures, ordinal data

sign test positive of negative 2 measures

CHI2 test if frequency of observation is different from some reference frequency, between groups

correlation variables affect each other

positive correlation both increase

quantitative analysis checklist 1. interpret which data levels 2. descriptives 3. identify test, h c 4. apply tests

purpose of academic knowledge contribute to scientific knowledge

academic paper structure 1. introduction 2. method 3. result 4. discussion 5. conclusion

good scientific practices correct sample, reference, grounding related work

scientific evaluation bias, reliability, validity

Klicka

Skriv

Lyssna

Spel

Skriv ut