Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
32 Cards in this Set
- Front
- Back
Descriptive |
Uses quantitative research methods to describe what is in order to gain an understanding of conditions that currently exist and the relationship between existing variables is not manipulated. May or may not involve hypothesis testing. Explore state of knowledge, small scale ,shorter. Describe a hypothesis (small or large) Analyze (small or large) |
|
Qualitative |
Variety of methods to explore existing phenomenon in a natural environment that generally yield non quantitative info in order to describe what is. Typically do not state a hypothesis before, but after. |
|
Research hypothesis types |
Directional - Posited when the researcher has reason to believe a certain relationship exists. Non-directional - Posited when researcher has no reason to believe a relationship/difference exists in any direction. |
|
Null hypothesis |
Research hypothesis not directly tested by data, it is the null hypothesis that states that there is no dif between groups/relationship between variables. |
|
Data collecting methods |
Observation, measurement, questioning |
|
Observation |
Direct - participants know they are being observed and why. Reactivity (when participant may act different when researcher present) is a concern Participant observation - researcher participates in activities as people being observed. |
|
Scaling techniques |
Measure the degree to which the research participant values or exhibits the desired construct. Eg like how you rate a class. Likhert scale - respondents are presented with a series of statements and asked to indicate the degree to which they agree/disagree. Semantic differential scale - use a continuum consisting of bipolar adjectives. Rating scale - Impression of behavior, can be numerical, or verbal, or rating importance. Structured alternative scale - A unique question and response format designed to reduce the tendency to provide a socially desirable answer Nominal, ordinal(frequencies and percentages only on the data) interval, ratio Controversy about the rating scales data due to the presumption that intervals between score points are equal. |
|
Common types of descriptive research |
Survey, developmental (longitudinal or cross sectional), case study, correlation, normative, observational, action, causal comparative. |
|
Structured questionnaire |
Questions with prescribed response alternatives that the participants need to choose. (Y/N, T/F, multiple choice) |
|
Unstructured Questionnaire |
Includes questions but no answers to them. |
|
Structured Interview |
Questions asked in order, no repeats permitted. Less bias as there is no variation from question script, interviewer does not need to know much about the subject. |
|
Unstructured interview |
Open ended questions, can answer freely. Has guide but not tied to it. |
|
3 characteristics of data needed to be worthwhile - 1. Objectivity |
The degree to which multiple scorers agree on the values of collected measured or scores (rater reliability). (free of tester bias) |
|
3 characteristics of data needed to be worthwhile - 2. Reliability |
Degree to which a measure is consistent. (measures consistently) |
|
3 Characteristics of data needed to be worthwhile - 3. Validity |
Degree to which interpretations of test scores or measures derived from a measuring instrument lead to correct conclusions. (Truthfulness, measures what it is supposed to measure) |
|
Sampling |
Use a sample from within a sampling frame, in order to make inference about the population based on the data. |
|
Who do you want to generalize to? |
The theoretical population. |
|
What population do you have access to? |
The study population |
|
How can you get access to them? |
Sampling frame |
|
Who is in your study? |
Sample |
|
Sample selection types |
Random, stratified random (select from within groups), Cluster (select the groups), systematic |
|
Cross sectional |
Only one observation of each unit of the study, not related to study length. |
|
Longitudinal |
Two or more observation sets collected for each unit of the study eg. follow ups. Expensive and time consuming, resource extensive. Threats to internal validity - instrumentation and mortality. |
|
Focus groups |
Group of participants interviewed together. |
|
Case studies |
One or more cases are studied individually. |
|
Correlation studies |
Determine relationship between two or more variables. Observations need to arise from the same source. Can be used to check instrument for validity and reliability. -1 = Perfect Negative relationship 0 = no relationship 1 = perfect positive relationship 0.90-1.00 = Very high correlation Strong relationship 0.70 -0.90 = High Correlation, Marked Relationship 0.40-0.70 = Moderate Correlation, Substantial Relationship 0.20-0.40 = Low Correlation,A relationship but its weak Less than 0.20 = Slight correlation,Relationship so small as to be negligible |
|
Continuous scores |
Have a continuum of scores. Can be decimals. |
|
Discrete scores |
Scores only possible with whole units. |
|
Types of descriptive statistics |
Measures of central tendency - mean, median, mode Measures of variability - standard deviation, range, confidence intervals. |
|
Correlation techniques |
'r' - used to designate the pearson product Moment correlation - requires interval or ratio scores, every participant has scores on two variables, most frequently used. |
|
Spearman rho or rank order correlation |
nonparametric technique for use with ordinal scores. Every subject has scores on two variables. |
|
Meaningfulness of 'r' |
Need 'n', the number of pairs of data used. Also need r squared, with indicates the shared variance between the two variables. |