1 down, more to go. Hopefully, everything will run smoothly.
WE CAN DO IT!
Regards,
Wani Nurfahani.
Sunday, 8 April 2012
-REFLECTION-
It was a long day but every minute is worthwhile. Thanks to every group who presented and not to forget, Prof Teoh for your patience in showing the path. :)
1 down, more to go. Hopefully, everything will run smoothly.
WE CAN DO IT!
Regards,
Wani Nurfahani.
1 down, more to go. Hopefully, everything will run smoothly.
WE CAN DO IT!
Regards,
Wani Nurfahani.
-QUANTITATIVE RESEARCH-
Quantitative Research (Part 5)
Survey Research
1. Researcher asks a series of question pertaining to the areas of interest in order to find answer.
2. 3 major characteristics
>Information
>Asking questions
>Data collection from sample
3. Major purpose
>to describe the characteristics of population
Types of Survey
1. Cross-Sectional Surveys
>information is collected at one point in time and from the sample that has been drawn from the predetermined population
2. Longitudinal Surveys
>since it studies changes over time, information is collected at different point of time.
>3 designs: trend study, cohort study, panel study
Steps in Survey Research
>defining the problem
>Identifying the target population
>choosing the mode of data collection
>direct administration to a group, web-based survey, mail survey, telephone survey, personal interview
>selecting the samples
>preparing the instrument
>types of questions: open-ended, closed-ended
>interview schedule
>suggestions on improving closed-ended questions:
>Be sure the question is unambiguous
>keep the focus simple
>keep the questions short etc
>pretesting the questionnaires
>ensure the overall format of the questionnaires
>Preparing the cover letter
>training interviewers
>using an interview to measure ability
Nonresponse
>This is referred to those members in a sample who do not respond.
>Conditions:
>Total nonresponse-the interviewees refused to be interviewed
>Item nonresponse-the respondents do not provide the answer to certain
questions.
>Suggestions on increasing the response rate feature the area of;
>administration
>format
-QUANTITATIVE RESEARCH-
Quantitative Research (Part 4)
Causal-Comparative Research
>to find out the cause or consequences of differences that already exist between or among group of individuals
>the group difference variable is either:
>a variable that cannot be manipulated (ethnicity) or
>a variable that cannot be manipulated due to several reasons (teaching style)
>disadvantages:
>lack of control (threats)
Steps Involved in Causal-Comparative Research
>problem formulation
>sample selection
>instrumentation
>design
>the basic causal-comparative design
Threats to Internal Validity in Causal-Comparative Research
>subject characteristics
>procedures:
a. Matching of Subjects
b. finding or creating homogeneous subgroups
c. statistical matching
Evaluating Threats to Internal Validity in Causal-Comparative Studies
>Subject characteristics
>SES, gender, ethnicity, marketable job skills
>Location
>Instrumentation
>Instrument Decay
>Data Collector Characteristics
>Data Collector Bias
>Other treats
>implementation, history, maturation, attitudinal and regression
-QUANTITATIVE RESEARCH-
Quantitative Research (part 3)
Correlational Research
>Study on the relationship between two variables; whether they are related or not.
The Purposes
>to help explain important human behaviours
>explanatory studies
>to predict likely outcome
>prediction studies
Basic Steps in Correlational Research
>problem selection
>sample selection
>instruments
>design and procedures
>data collection
>data analysis and interpretation
Threats to Internal Validity in Correlational Research
>Subject characteristics
>Location
>Instrumentation
>Instrument Decay
>Data Collector Characteristics
>Data Collector Bias
>Testing
>Mortality
-QUANTITATIVE RESEARCH-
Quantitative Research (Part 2)
Single-Subject Research
Characteristics
>data are collected and analyzed for one subject at once.
>to study changes in behaviour of an individual after treatment is being given.
Single-Subject Designs
>the graphing of single-subject design
>line graphs
>dependent variable: vertical axis
>a sequence of time: horizontal axis
>the first: baseline
>the second: condition line
1. The A-B Design
>A-Baseline, B-Condition line
>the researcher does not know whether the change of behaviour is due to the treatment or not.
2. The A-B-A Design
>also known as reversal design
>A-Baseline, B-Condition line, A-Baseline
>present stronger evidence
>ethical disadvantage: leaving the subject to the A condition.
3. The A-B-A-B Design
>A-Baseline, B-Condition line, A-Baseline, B-Condition line
>increase the effectiveness as both conditions are tested twice.
>the ethical disadvantage is avoided.
4. The B-A-B Design
> B-Condition line, A-Baseline, B-Condition line
>to be implemented when the behaviour of subject is disturbing to the point where the baseline could not be established
5. The A-B-C-B Design
>A-Baseline, B-Condition line, C-Contigent Condition line, B-Condition line
6. Multiple-Baseline Designs
>An alternative to the A-B-A-B Design
>to be conducted when it is impossible to create baseline or condition line
>the researcher determines the baseline and the condition during the process.
-QUANTITATIVE RESEARCH-
Quantitative Research (Part 1)
Experimental Research
The uniqueness
1. purposely conducted to influence a specific variable
2. able to test the hypotheses of a research on cause-and-effect relationships.
Essential characteristics
1.Comparison of groups
>Consist of;
>Experimental group – with treatment
>Control group – without treatment or
>Comparison group – with different treatment
>The comparison ensures whether the treatment is effective or not.
2.Manipulation of the independent variable
>The researcher determines the treatment and which group will be the experimental group.
>The independent variable is tested in a number of ways;
>One variable vs another
>Presence vs absence of treatment
>Varying degrees of the same treatment
3.Randomization
>It features;
>Random assignment:
every sample has an equal opportunity of being assigned to either experimental or control group.
>Random selection:
every member of the population has an equal opportunity of being assigned to be a member of the sample.
>Three essential notes
>This is conducted before the experiment starts.
>This is a process of assigning individuals to groups.
>This enables groups that are equivalent to be formed by the researcher.
>Random assignment is meant to discard extraneous variables.
Control of Extraneous Variables
1.This is crucial as to discard/minimize the effect of extraneous variables.
2.Some of the suggested ways;
>Randomization
>Holding certain variables constant
>Building the variable into the design
>Matching
>Using subjects as their own controls
>Using analysis of covariance
Group Designs in Experimental Research
Considerable as Poor or Weak Experimental Designs;
1.The One-Shot Case Study
>no comparison = no control
2The One-Group Pretest-Posttest Design
>No comparison=the researcher would not know whether the result is due to the treatment or not
3.The Static-Group Comparison Design
>Also known as nonequivalent control group design
>Static means two existing groups are used. Therefore, it is not a random assignment.
>This may lead to extraneous variable = differential subject characteristics
4.The Static-Group Pretest-Posttest Design
>Pretest is given to both groups.
>Posttest–Pretest = Gain/Change
>Gain/change=improve (depends on the cases)
True Experimental Designs
1.The Randomized Posttest-Only Control Group Design
>features independent variable and control groups that are formed by random assignment
>the presence of control=good
>the presence of threats: mortality
2.The Randomized Pretest-posttest Control Group Design
>the presence of pretest
> features independent variable and control groups that are being tested twice (pretest and posttest)
>pretest may lead to pretest treatment interaction threat
3.The Randomized Solomon Four-Group Design
>to discard the possible effect of pretest
>features 4 groups;
a.2 groups with pretest
b.2 groups without pretest
>enables the best control of the threats
>drawbacks;
a.requires large sample
b.requires more energy and effort
4. Random Assignment with Matching
>pairs of individuals could possibly be matched on certain variables
>this could be done in two ways:
a.Mechanical Matching
>a process of pairing two individuals with similar scores on a specific variables
>it is difficult as it requires a large sample and there are samples that will be taken out from the research as they could not be paired with anyone.
b.Statistical Matching
>each sample is given predicted scores
>the predicted scores will be compared with the actual scores
>the difference is then used to pair the samples
-QUALITATIVE RESEARCH-
Qualitative Research
Content Analysis
>A technique that allows researcher to study human behaviour in an indirect way.
>This could be done through the analysis of human communication.
>widely used in educational research.
Categorization in Content Analysis
>The categories are determined in advanced.
>The categories becomes the analysis after the researcher gets familiar with the descriptive information
Steps Involved in Content Analysis
>Determine objectives
>Define terms
>Specify the unit analysis
>Locate relevant data
>Develop a rationale
>Develop a sampling plan
>Formulate coding categories
>Check reliability and validity
>Analyze data
Advantages
>Unobtrusive
>Useful
>Allows researcher to delve into records and documents
>Simple and economical
>Permits replication of a study
Disadvantages
>Limited to record information
>In establishing validity
Thursday, 5 April 2012
-OBSERVATION-
Observation
In collecting data for a research, it could be done through observation. The researcher could observe the situation of interest or how the samples act while it takes place. This is in fact genuine as the situation is witnessed before the researcher’s own eyes.
There are four terms that are commonly used to describe different degrees of roles from a complete participant to a complete observer;
- complete participant
- participant-as-observer
- observer-as-participant
- complete observer.
These terms could be found in the four types of observation available.
Types of Observation
I. Participant observation
- Participant observation is the situation where the researcher literally participates in the research study.
- The degrees of the researcher’s participation vary from one level to another.
- In the situation where the researcher chooses to be a complete participant, he should fully participate in the activities of the group without letting anyone in the group knows his identity as the researcher.
- It is the other way around if the researcher chooses participant-as-observer as his role in collecting data for the research. It is different in a sense that the researcher should make a clear understanding to the group that he is doing a research. However, similar to the role of a complete participant, he also needs to fully participate in the activities of the group that he studied on.
- The participant observation could be either overt or covert to the group that is being studied by the research. This is depended on the situation or the type of study that the researcher involved in.
- Non-participant observation is the type where the researcher does not involve literally in the situation he is observing.
- There are two types of roles of the researcher which are;
- In the situation where the researcher chooses the role of observer-as-participant, the researcher should reveal his identity as the researcher and involve in the activity superficially.
- It is totally a different story if the researcher chooses the role of a complete observer. The researcher just sits and watches the activities.
III. Naturalistic observation
- Naturalistic observation features observations in natural settings.
- In this situation, the researcher does not participate in the activities and just observes and records the situation as it is.
IV. Simulation
- The kind of observation where the researcher asks the samples to act out a situation and records the role-play.
- It is either in a form of individual or team role-playing.
- There is a major disadvantage of simulation as a means of collecting data which is the artificiality. The researcher could not guarantee that the role-play is the real thing happening in the actual situation.
Observer Effect
- Logically, observation affects those who are aware that they are being observed. Therefore, the samples might not behave naturally during the observation. This is called as observer effect. In one way or another, it might affect the observational data of the study as well.
- However, this could be catered if the researcher is willing to spend more time to stay with the samples. Perhaps, the samples will act naturally while they are being observed as they are used to the researcher being around them.
Observer Bias and Observer Expectation
There are issues that put observation at concern. There are observer bias and expectation.
1. Observer bias
The situation where the observational data of the research is influenced by the researcher’s past experience or beliefs. Therefore, the researchers practise a number of ways in order to avoid this issue. As for the qualitative researchers,
- they study the subjective factors objectively in a team so they can check on each other.
- they spend some time to get to know the samples and the settings before they collect the data.
- The data is collected from a various perspectives using different formats.
- They also reflect on their subjectivity as a part of their field notes.
2. Observation expectation.
Usually, the researchers already know the target subject that they want to observe. Therefore, they will have certain expectations of how the samples should behave or react. To their surprise, sometimes the samples are not like what they are expecting. In this case, audiotaping and videotaping are important in order for the researcher to monitor the observation closely.
Coding observational data
- A coding scheme is used primarily in quantitative research to observe the frequency of the samples’ behaviour.
- The researcher judges and codes the behaviour into categories as it occurs.
The Use of Technology
- The use of technology could always be referred to the videotaping or audiotaping of a research.
- The researcher could play it repeatedly later, after the observation and manage to see the overlooked part during the observation and determine the category of the behaviour in a more casual and relaxing moment.
-CHECKLISTS-
Data-collection Instruments
In administering the data-collection instruments, it could be done by the researchers themselves or by the subjects.
Researcher-completed instruments
Performance checklists are to be completed by the researcher.
- A performance-checklist features a list of behaviours.
- The behaviours usually portray a certain type of performance. The real purpose of performance-checklist is to find out whether the subject is behaving in a desired way when he or she is completing the task or not. This could be seen through the number of check marks that the subject has drawn on the checklist.
- In addition, the rating scales will be used in a checklist if the researcher has the intention to know how well the process is when the subject performs the task.
- The example is on page 122, figure 7.11
Subject-completed Instruments
On the other hand, self-checklists are to be completed by the subjects.
- A self-checklist illustrates a list of criteria related to the subjects of a study.
- The criteria could be the characteristics, activities and many more.
- If the subject met the characteristic or perhaps performed the activities, the subject should mark the particular criteria in the list.
- the example is on page 125, figure 7.13
The Purposes
- As a guide to build skills in writing research proposals
- To develop the quality of research proposals
- To guide evaluative feedback
- To provide formative evaluation thus supporting the summative evaluation
- As a guide to evaluate previous proposals
Lynnda J. Emery, Carolyn Harvey and Catherine M. Andersen. (2006). Formative Evaluation Using Checklists to Improve Research Proposals.
-INTERVIEWING-
Interviewing
- Interviewing allows the researcher to check the accuracy of data collection.
- The purpose of conducting the interviews is to discover the views, opinions and thought of the samples as well as the things that the researcher often overlooked during the observation.
Recording data
- Recording data involves audio tapping, videotaping and also taking notes.
- This is essential to keep track the conscious and unconscious response of the respondents.
Ethics in interviewing: the necessity for informed consent.
- It is preferable to request participants to sign an informed consent form.
- This is to preserve the rights of participants and avoid misunderstanding of any issues regarding the interview.
There are four types of interviews that can be used in a qualitative research; structured, semi-structured, informal and retrospective.
- Structured and semi-structured interviews are conducted in a formal manner. In collecting data, the researcher asks a series of questions and later, the information will be compared and contrasted. This is effective in testing the hypothesis of the research.
- Informal interviews are conducted in a casual manner. There are no certain guidelines of asking the questions. The objective of this interview is to gain the views, thoughts and opinions of the samples.
- Retrospective interviews could be conducted in various forms; structured, semi-structured or informal. However, this is the least favourable type of the interview as it requires the samples to recall and reconstruct of something that happened in the past. Therefore, there is a possibility that the data is not accurate thus unreliable for the research.
Four major interview strategies
There are four major interview strategies; informal conversational interview, interview guide approach, standardized open-ended interview and closed, fixed-response interview.
- The characteristic of informal conversational interview is the questions are not planned and they are asked in accordance to the situation. Therefore, the questions can fit in the samples. However, this leads to vast different responses from one sample to another thus the data could not be easy to analyse.
- In interview guide approach, the researcher has outlined the topics and issues to be asked. As for the questions and sequence, it will be determined during the interview. The outline offers a path for data analysis thus make the data easier to be analysed. However, since the responses are all due to the perspectives of the samples, the responses of the samples might be different from each other. Thus, the data could be difficult for the researcher to analyse.
- In practising standardized open-ended interview, the researcher has planned the questions and its sequence before the interview is actually conducted. Therefore, all the samples are asked the same questions by the researcher. As the samples are asked the same questions, it increases the comparability of the responses. However, it limits the naturalness and relevance of the questions to the samples.
- Before closed, fixed-response interview is actually conducted; the categories for the questions and responses are already prepared by the researcher. Although it limits the naturalness of the answers of the samples, the data becomes simpler and clearer to be analysed and compared.
Key actor interviews
The term ‘key actor’ derives from the term ‘key informant’.
- Key informant is the individuals who know their culture and history and are able to articulate better than others.
- Key actor is almost similar to key informant. Key actor is a knowledgeable and informative individual. In a research context, researchers should take some time to identify and build rapport with the key actors of a population.
Types of interview questions
There are six types of basic questions which are; background, knowledge, experience, opinion, feeling and sensory questions.
- Background questions feature questions related to respondents such as education level, age and many more.
- Knowledge questions refer to the facts that the respondents know such as the information of the school and many others.
- Experience questions revolve around the previous and current behaviour of the respondents.
- Opinion questions concern on the respondents’ views, opinions and thoughts.
- Feeling questions are related to how respondents feel on certain things.
- Sensory questions rely on the five senses. This is solely related to what is seen, heard, tasted, smelled and touched by the respondents.
Interviewing behaviour
There are a number of expectations of how the researchers should behave during the interview. The researcher should;
- be natural
- build rapport with the respondents
- respect the respondents and their culture etc
Focus group interviews
- A group of small people who is seated together and being asked a series of questions to think of.
- The objective is to identify what this group of people think of the questions asked.
Subscribe to:
Posts (Atom)