Ethnomethodology, Document study and Reliability (Notes)
Advantage of Ethnomethodology
a) It study both verbal and nonverbal behavior
b) Its longitudinal because it is ongoing and changes of behavior can be viewed over time
c) It gives insight onto what and people think about common place activities and behaviors why
Disadvantages
a) It investigate the process of how behavior occur rather than the product of that occurrence
b) You can not investigate large scale studies but smaller investigations
Document study
Is reading documents of the institution you are studying, eg. Factors for effective guidance and counseling services in secondary schools in Tanzania
i. Routine client records
ii. Correspondence from and to staff
iii. Financial charts
iv. Official and unofficial documents generated by or for the program
The above documents provide the researcher with basic sources of information regarding activities and processes of the schools, and can view other questions not previously considered to follow up on observations participants observation or ethnomethodological research.
Types of Documents
a) Internal documents e.g. Memos in large institutions like UDOM it's the amount of papers that flow from top to bottom. Of course some flow in the opposite direction as well
b) External communication is the materials that circulate outside the organization
c) Personal records is about hiring and firing practices, promotion and rewards
NB: A good qualitative study will most likely put to use most of these methods as they are all beneficial.
Validity and Reliability
Validity in research is how sound your research is.
You can consider validity in four types, these are
a) Internal Validity is attained when the researcher managed to control all the extraneous variables.
e.g.. In a study of the effectiveness of teaching method, lets say mediated learning experience (MLE). One group will be taught using MLE while the other will not.
After sometime give the same test to both groups.
If you managed to control all extraneous variables, then you will attain internal validity. Internal validity is very difficulty to achieve.
As you try to control extraneous variables, external validity is threatened
So, you must compromise because you can not control all threats..
Threats of internal validity
Maturation due to time that has elapsed
Test retest can give a clue of the study
History events that occur simultaneous with the study
Instrumentation is the use of different tools can change results
b) External validity is the researchers` ability to generalize the findings of a study.
As a researcher when you control threats of internal validity you create artificial environment which generalization of other settings is impossible.
Threats of external validity
Reactive of testing
The interactive effects of selection of respondents
The reactive effects of experimental arrangement
Multiple treatment interference
c) Statistical Validity is referred to the accuracy of the conclusion drawn from the statistical test.
Eg. If your results show that enrolment at UDOM is 70% males and 30% females, then you conclusion that there is equal opportunity at UDOM.
Such conclusion is wrong
d) Construct validity address the degree to which the underlying theory of research explains the observed results.
E.g. Theory show that adolescents are sexually active as such they are vulnerable to consequences such as unplanned pregnancies, induced abortion, STDs including HIV/AIDS
However, your findings from 1000 students shows that 12-20 years students none is sexually active, then these results are not valid.
Reliability
We often speak of reliable friend
We often speak of reliable machine
News, people speak of reliable source of information
Whatever the case may be, what is reliability?
Reliability
A measure is considered reliability if it would give you the same results over and over again (when you assume what you are measuring is not changing) is a composition of two components: true ability (or true level) of respondent on that item; and random error
For example, You observed Juma`s score of 85% in SE 301 timed test.
However, in reality the ability of Juma might be better than the score indicated eg. 89%
So the error for Juma is 89%-85%= -4%
What does this mean?
While Juma`s ability is 89% he may have had bad day, like no breakfast, quarreled with someone, some distract while doing the test etc.
Measures like these can contribute to errors in measurement that makes the students' observed abilities appear lower than their true or actual abilities.
If your measure of 85% is reliable, you should find that if you measure or observe twice on the same persons the score should be pretty much the same
Why would they be the same? How do you know?
The two error scores(e1 and e2) have different values, because you are likely to have different errors in different occasions.
However, the true score “T” is the same for both tests
NB: Sometimes errors will lead Juma to perform better on test than his ability eg. good day for guess work. The error is randomly distributed
Types Reliability
i. Inter rater or inter observer reliability is used to assess the degree to which different raters or observers give consistent estimates of the same behavior.
ii. Test retest reliability is used to assess the consistency of a measure from one time to another.
iii. Parallel forms reliability is used to assess the consistency of the results from one test constructed in the same way from the same content domain.
iv. internal consistency reliability is used to assess the consistency of results across items within a test.
When you are done with validity and reliability you move onto data analysis plan
Data analysis plan
What is data?
Is something known or assumed as fact
A thing given as a basis of reasoning or calculation
What is variable?
Variable is observed quantity or attribute which varies from one member of the population or sample being studies to another.
Dependent variables
Dependent or response variable (s) are those depend upon the values of independent variables. Commonly used symbol is Y
For example
1. Students performance
2. Teachers motivation
3. Knowledge, attitude and practice
Independent variables
Independent or explanatory variable is the one that is manipulated by the researcher; Commonly use X symbol
For example
1. Working hours
2. Age
3. Gender
4. Years of experience
5. Level of education
Measurement scales
a) Nominal scale is mutually exclusive category that varies qualitatively eg. Gender, religion, region
For computational purposes, numbers are attached to the categories when using SPSS
E.g. Gender: man=1; woman=0
Eye colour:1=blue, 2=brown, 3=black
Note that these doesn't mean the categories are unequal
b) Ordinal scale is more informative than nominal and differ from nominal in that the categories are preserved in the analysis. For computation, numbers are attached to reflect their relative order.
Social economic status: High=3; medium=2; low=1 or 1st, 2nd, 3rd etc.
Note that the difference between medium and high is not equal as between low and medium. Even though numerically 3-2=1and 2-1=1
The numbers just give the rank.
c) Interval scale is more informative than nominal and ordinal scales. The difference between adjacent categories are equal, however, there is no true ZERO point.
E.g. Centigrade temperature scales where 40oC is higher than 30oC and the difference between 30oC and 40oC is the same as 10oC and 20oC
However 40oC is not twice the temperature of 20oC, since 0oC is not absolute ZERO.
d) Ratio scale differs from interval scale that there is true zero point. Ratio scale is more informative that the 3 preceding scales
E.g. Consider the height in meters. A stick 6 meters. A stick 6 meters long is twice as much as 3 meters long stick or a baby 12kg weight is twice as much heavier than a 6kg baby.
Post a Comment