Reading...
Play button
Play button
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;
47 Cards in this Set
- Front
- Back
|
Manifest coding
|
- explicit, stated content
-quantitative and reliable measure - coding by counting the frequency, a word occurs or space for image or article - coding by categorizing the intents or directions of connation of words |
|
|
Latent Coding
|
- implicit, underlying content
-qualitative and valid measure - coding based on the characterizing implied message -coding based on subjective scores to rate the degree of message or image |
|
|
Strengths of Context Analysis
|
-flexible
- easy to replicate - cheaper -easy to access |
|
|
Weakness of Context Analysis
|
- media must reoccurred, to analysis
- social artifacts may be hard to review |
|
|
Secondary Data Analysis
|
-analyze data by others, means
example Durkheim study of suicide |
|
|
Threats to validity and reliability
|
- variables may not met your content
- units may not meet your level of threats - documentation may be limited |
|
|
Ways to improve validity
|
- use logic reasoning from Durktheim theory
- use replication via multiple measures |
|
|
Ways to improve reliability
|
- Anticipated, likely biases
- Logical reasoning from theory - Replication with multiple measures |
|
|
Strength of SDA
|
- cheap
- data already collected - broadly represented - standardized - useful for comparison |
|
|
Weakness of SDA
|
- quantitative, not qualitative
- face threats in reliability and validity |
|
|
Elements of Historical Comparative Research
|
1. Evidence is often incomplete, fragmentary and suspect
- reliance of trigulation and use of mutiple use of data 2. Immersion in data, in order to understand data - to improve the validity of intrepretation |
|
|
Design in HCR
|
- Comparative analysis is a broad term
- Data sources for comparative analysis are usually secondary. - International agencies and publications - Cross national and health surveys |
|
|
Issues with sources of historical data
|
Primary sources- original resources (media &natural good)
- incomplete, hard to locate biases - critically evaluate authors credentials and claims Secondary sources- histories, other products of prior research - histographic approach to selection if evidence - identify author commintment, be careful for underlying factors |
|
|
Types of Qualitative Management
|
Transcription of Files
Types of Files Manual and automatic coding Coding procedure for grounded theory |
|
|
Qualitative Data Mangement
Transcription to files |
transcribe field reports by typing them up
easy to read, allow for back up |
|
|
Qualitative Data Mangement
Types of files |
Master files- all files in chronological order
Additional file are by type of notes, event and information |
|
|
Qualitative Data Mangement
Manual and Automatic coding |
Manual- read through, highlight key words and phrases
Auto- conduct word searches using computer program |
|
|
Qualitative Data Mangement
Coding procedure in grounded theory |
Open coding- identify key concepts and categories
Axial coding- identify association of the different concepts - find precursors and consequences Concept mapping- graphical display of concepts and their relationships Memoir- comment on codes, coding and omergent theory Operational notes- comments on data |
|
|
Qualitative Analytical techniques
|
1. Idealization n
2.Successive approximation 3.Illustrate method 4.Multiple sorting procedure 5.Qualitative comparative analysis 6.Domain analysis 7.Event structure analysis |
|
|
Qualitative Analytical techniques
Idealization |
Idealization- classic method developed by Mack Weber
a) Ideal type- pure and essential form of a case b) Specially describes the essential traits of ideal type. |
|
|
Qualitative Analytical techniques
Successive approximation |
Successive approximation- an extension of idealization
-To modify the ideal, collect more data and compare ideal to data |
|
|
Qualitative Analytical techniques
Illustrate method |
a) Is deductive and works from a theoretical framework
b) Must identify concepts and relationship lines |
|
|
Qualitative Analytical techniques
Multiple sorting procedure |
An elaboration on contrast question in independent intervals
- give interviewees a series of cards relevant to topic - interviewees sort card in pile - then use context question to understand the criteria for the sorting concept |
|
|
Qualitative Analytical techniques
Qualitative comparative analysis |
- a inductive comparative method
a) Method of agreement- identifying characteristics that occur most often b) Method of difference- identifies difference in other similar cases and relates events. |
|
|
Qualitative Analytical techniques
Domain analysis |
an organized ideal induced from a text
- defined by key somatic relationships among key terms of data |
|
|
Qualitative Analytical techniques
Event structure |
a chronological sequence of causes
|
|
|
Quantitative data management
|
Strategies for quantifying data
Possible Code Cleaning Contingency Cleaning Collapsing categories |
|
|
Strategies for quantifying data
|
Deduction and Induction
|
|
|
Strategies for quantifying data
|
Deduction: adopt codes that reflect theoretical distinctions
-categories need to reflect a key theoretical distinction - examples: Class distinctions, family types ethnic differences, etc. Induction:identify distinctions from patterns in observations -code all different responses in distinct categories -then take the popular categories & collapse smaller but similar categories into them - example: Cults |
|
|
Possible Code Cleaning
|
code things into categories that you defined.
- if gender is 1&2, there shouldnt be a 3 - easy to check for frequency distribution |
|
|
Contingency Cleaning
|
response to a Q must be consistent with contignent Q’s
Steps: -Select one code for variable -Run frequency distribution on contingent variables -Look for inconsistent cases -Then select next code for first variable and repeat. |
|
|
Collapsing categories
|
Why do it: Allows for finer distinctions
When NOT to: i. If you only have 2-3 cats ii. Cases are nearly distributed among categories iii. Collapsing interferes with theoretical distinctions |
|
|
Univariate analysis
Measures of central tendency & dispersion |
1. Measures of central tendency –different kinds of averages
a) When appropriate for mean, median and mode i. Mean- for continuous variables ii. Median- for all but nomial variables iii. Mode- all level of measurement 2. Measures of Dispersion- distribution of values for cases around the central tendency b) When interpretation for frequency distr, range, and standard dev. i. Frequency distri- easy to visualize ii. Range- Best from continuous variable iii. Standard Dev- Appropriate for continuous variable |
|
|
Multivariate descriptive stats
|
Univariate descriptive stats
Multivariate prescriptive stats Measures of Association Regression |
|
|
Univariate descriptive stats
|
summarizes large sum of data
-measure central tendency and dispersion |
|
|
Multivariate descriptive stats
|
a quantity relationship
-measure association and regression analysis |
|
|
Measures of Association
Lamba (A) |
i. For nomial DV’S and Categorical DV’s
ii. Lamba ranges from 0 (no association) to 1 (perfect association) iii. Example: Gender and employment= if cases differs from each other than there is an association= 0.67= strong association |
|
|
Measures of Association
Gamma (Y) |
i. For two ordinals variable; measure association between ranks
ii. Ranges from -1 ot 1; indicate strength and reaction of association. example:- If high ranks of the IV correspond to high ranks of DV Gamma, then positive 0 to +1 |
|
|
Measures of Association
Pearson’s R or Correlation Coefficent |
i. For continuous variable, measure the association between values
ii. Ranges from -1 to +1; indicates strength and direction of association |
|
|
Regression
|
Interpretation of B
Intrepretation of R Multivariate Regression |
|
|
Interpretation of B
|
estimates the slope coefficient B
ii. B describes, a line that best fits the scatter plot of IV and DV iii. Value of B qualifies the effect of an IV and the DV iv. B is mathematically related to r, but the two are not the same |
|
|
Intrepretation of R
|
i. Measures the strength of association
ii. Ranges from 0 (no explanation) to 1 (completed explanation) iii. The higher the R, the higher the B, and the higher the R2 |
|
|
Multivariate Regression
|
powerful; allows for multiple IV and estimates of B for each
|
|
|
Purposes of interential stats
|
a) Limitation of descriptive stats
i. Do not test hypothesis or draw conclusion ii. Do not allow generalization in pop b) Purposes of infertinal stats -Provide a basis for testing hypothesis and draws conclusion -Generalized |
|
|
Test Statitics (when appropriate)
|
c) Test the significance difference in relationships of varaiables
d) Large group difference means-large test stat |
|
|
P-Values
|
a) In regards to t-test: qualify the staticial significance
b) In regards to hypothesis testing- finding is due to sample error |
|
|
Bivariate and Mulivariate Inference
|
a) Logic, compare the test stats ot the criticial value
b) Test stat value reflect group differences c) Steps i. Choose test stat ii. Adopt level of significance iii. Calculate test stat and find p-value iv. Compare test stat to criticial value |
