Inductive And Qualitative Approach Versus Deductive And Quantitative Education Essay

The study is descriptive nature. ‘Research philosophy is an over-arching term relating to the development of knowledge and the nature of that knowledge’ Saunders et al, (2009). Since, the research is guided primarily by the ‘scientific criteria’ of the measuring instruments of quantification, systematic collection of evidence, reliability and transparency, researcher adopted ‘positivism’.

3.2. Research strategy

3.2.1. Inductive and qualitative approach Vs Deductive and Quantitative study

Research on special education was vast and thus in order to further strengthen the findings researcher adopted quantitative deductive approach where theories are tested through empirically. Data was collected through pre-determined instrument to obtain numerical data which can be analyzed statistically.

3.3. Study setting and Sampling method

The research project took place at the randomly selected schools in the United States. The teachers in these schools were interviewed on their perceptions and attitude towards special education program. About 200 teachers who are trained under special education (previous experience with certificate in special education) and not trained will be selected using systematic sampling method. The study adopts a pure descriptive approach. Data on the demographic information of the study sample were done based on the following criteria: whether the instructors included in the study was married or single, whether they were professionally trained for special education or not, the experience of the instructor greater than or less than 10 years.

Predictor Variables

It is vital that the participants fill in a detailed biographical questionnaire that gives information on the gender, marital status, experience as these demographics are predictor variables on the attitude towards inclusion.

3.4. Pilot study

In order ensure for the content, readability and ambiguity the pilot study will be conducted prior to the main study. Pilot interviews were carried out among a small group of teachers, to generate items for the scale in assessing the attitudes of teachers towards the inclusion of special needs children in general education classrooms. The final scale consisted of 20 items which were accompanied by five-point Likert-type self-report rating scales ranging from “positive attitude” to “negative attitude” (1 to 5).

Procedure of Data Administration

The researcher administered the instrument in each of the selected schools after obtaining their mission to do so from the school authorities. In each of the schools, respondents were gathered in a class and were administered the questionnaire. The instructions were read to the respondents as regard the filling of the questionnaire. The items in the questionnaire were properly filled and returned after the exercise. To ensure there was no case of any loss of items as return rate was assessed.

Snow (1974) recommended eight conditions to make designs more representative:

1. Actual educational setting: This survey was distributed in the actual educational setting of the teacher participants.

2. Variation of the educational setting: The four schools were chosen primarily because they were geographically and socio-economically varied. One upper middle-class and one lower middle-class school was examined in each county.

3. Observation of the participants: The researcher observed (a) all teacher participants during the pilot portion of the survey development and (b) the teacher participants surveyed during the study.

4. Observation of the social context: The researcher made a minimum of three visits per school to observe the social context.

5. Preparation of the participants: Brief instructions were given in the cover letter, on the survey, (and in person, for the pilot portion). Strict protocol and procedures were followed. Treatment fidelity was observed.

6. Incorporation of a control treatment that uses customary approaches: The survey was designed to be understood and completed simply, using common pen-and-paper assessment techniques.

3.5. Time Horizon

Cross-sectional technique was adopted where data collected at one point of time and due to its inexpensive to conduct.

3.6. Data collection

3.6.1. Primary data collection

This descriptive study involved mainstream classroom teachers and special education teachers. Questionnaire method was used to collect primary data. Questionnaire was developed based on the following hypotheses: Four hypotheses were postulated at the significant level of .05; they are:

H01: There is no significant difference between male and female teachers in their attitude towards the inclusion of special needs students in general education classrooms.

H02: There is no significant difference between married and single teachers in their attitude towards the inclusion of special needs students in general education classrooms.

H03: There is no significant difference between professional qualified and non-professional qualified teachers in their attitude towards the inclusion of special needs and children in general education classrooms.

H04: There is no significant difference between teacher with less than 10 years of teaching experience and their counterparts with more than 10 years of teaching in their attitude towards the inclusion of special needs students in general education classrooms.

3.6.2. Secondary data collection

A desk-based approach was also adopted for the research where the data in collected from academic publications, journals, news-papers, government publications, policies, annual reports, and company websites.

3.6.1.1. Research instrument

A survey on the attitudes and knowledge of school teachers regarding inclusive education was conducted. It consisted of an 18-item scale, divided in three parts: a) teachers perceptions (8 items), assessment of teachers’ views with the claim that children with disabilities are entitled to education together with their typically developing peers in inclusive classrooms, b) collaboration between the mainstream and special education teachers (5 items), which explored the relationship between the mainstream and special education teacher and c) strategies to improve inclusive education (5 items), which examined how inclusion can be enhanced. The participants were asked to indicate their degree of agreement on a five-point Likert scale In order to complete the questionnaire (1 = Strongly Accept; 2 = Agree; 3 = Undecided/Neutral; 4 = Disagree; 5 = Strongly Reject).

Questionaire: Part I

Students with special needs fare better academically in inclusive education

Children with special needs must be integrated into the regular student community

Students with special needs must be placed in regular classes with back up support to achieve highest level of inclusion

Academically talented students may be isolated in inclusive class rooms

Placement of children with special needs in regular class rooms may negatively affect academic performance of mainstream students.

Children with special needs will benefit from inclusivity

Children with special needs have a right to receive mainstream education

Labelling as stupid, weird, hopeless is a problem in inclusive education.

Questionnaire: Part II

Special needs teachera and regular teachers need to work together in order to teach students with special needs in inclusive classrooms

Although the inclusive education in a concept, its implementation is ineffective due to objections from mainstream classroom teachers

Mainstream teachers have a main responsibility towards the students with special needs placed in their clssrooms

The presence of a special education teacher in the regular classrooms could raise difficulties in determining who really is responsible for the special students

The special education teacher only helps the students with special needs.

Questionnaire: Part III

Mainstream classroom teachers have the training and skills to teach special needs students

Special needs students need extra help and attention

Students with special needs committed more disciplinary problems compared to the regular students

Mainstream classroom teachers received little help from the special needs teachers

Although inclusive education is important, the resources for the students with special needs in a mainstream classroom are limited.

3.7. Reliability, validity issues

The reliability and validity of an instrument will be done through pilot study and face and content validity measures.

Validity

No matter what research design is selected, concern for factors that could affect the validity of the design is always primary. Typically, two types of validity are considered when designing research: (a) internal validity and (b) external validity. Although both types of validity are important, emphasis may vary depending on the type of research questions being investigated. For descriptive questions (as in this study), external validity receives greater emphasis because the priority of the researcher is to systematically investigate an existing sample of individuals or phenomenon, as opposed to studying the impacts of a phenomenon or intervention (as in experimental research). The factors jeopardizing external validity (or representativeness) are often more relevant to a descriptive study.

Internal Validity

Internal validity determines whether, in fact, the experimental treatments used made a difference in a specific experimental instance (Campbell & Stanley, 1966). Relevant to internal validity, Campbell and Stanley identified eight classes of extraneous variables, which, if not controlled by the experimental design, could produce effects confounded with the effect of the experimental stimulus. Cook, T. and Campbell (1979) expanded the list to include 12 extraneous variables. The variables and their relevance to the design of this study are reviewed below:

History:

History addresses the specific events that occur between the first and second measurement in addition to an experimental variable (Campbell & Stanley, 1966) and would only be a potentially relevant threat in this design in relation to the 15 teachers randomly selected for participation in the confirmation interview. Since these interviews were completed shortly after the survey participation, and are only used for confirmation purposes, the threat is minimal.

Maturation effects:

Maturation effects are defined as those processes (physical or psychological changes) within the participants that are operating as a function of the passage of time (Campbell & Stanley, 1966). Inherent within the research design was the use of only one treatment (the survey), which takes approximately 20 minutes to complete. The possibility is nominal that the growth of hunger, tiredness, or other conditions, within that time period would impact the data.

Testing effects:

Testing effects (defined by Campbell and Stanley [1966] as those effects of taking a test upon the scores of a second testing) were also controlled by this design–as only one test was used. The pilot participants were not used as study participants and the participants used for interviews were not reassessed–but were only asked to confirm their answers.

Instrumentation:

Instrumentation (Campbell & Stanley, 1966) refers to changes in the calibration of a measuring instrument, observers, or scorers used, and can produce changes in the obtained measurements. Controls built into this design for instrumentation effects included the use of one measurement (survey). The instrument was (a) carefully developed by accepted guidelines; (b) piloted; and (c) self-administered with supervision, handling, and mindful interpretation by only the researcher who had insight of the threat potentials. Experimenter bias and treatment fidelity were consciously avoided.

Statistical regression:

Statistical regression (explained by Campbell and Stanley [1966], as when groups have been selected on the basis of their extreme scores), was not considered a relevant threat in this design because only one test was applied, and selection was dependent upon general experience criteria and availability, not test scores.

Differential selection:

Biases, which result from differential selection by the comparison groups (Campbell & Stanley, 1966), were not viewed as a significant threat in this research design because no comparison groups were used. The design used was more descriptive in nature, and the purported generalization was limited to the teachers of the four assessed schools.

Experimental mortality:

Experimental mortality, or differential loss of respondents from the comparison groups (Campbell & Stanley, 1966), is controlled within the study design because no control groups were used, and the study was completed in a relatively short period of time. The possibility of the absence of some significant (main group) participants at the time of assessment is a noteworthy threat although deemed unavoidable. The researcher had no control over participants’ absences.

Selection-maturation interaction:

Selection-maturation interaction is where certain designs are threatened due to the given respondents growing older, or the results may be specific to the respondents’ given age level, fatigue level, etc. (Campbell & Stanley, 1966). These threats were not relevant to this design because, again, no pretest or comparison groups were used and the questionnaire was taken by various aged participants within a short period of time.

Experimental treatment diffusion, compensatory rivalry (John Henry effect):

Experimental treatment diffusion, compensatory rivalry (John Henry effect), is nominal compensatory equalization, and resentful demoralization. Experimental treatment diffusion, compensatory rivalry (John Henry effect), compensatory equalization, and resentful demoralization (Cook, T. & Campbell, 1979) as threatening extraneous variables were immaterial because no control group was used in this design.

External Validity

External validity (or representativeness) is the extent to which it is possible to generalize from the data and context of the research study to broader populations and settings (Bickman, 1989; Cook, T. & Campbell, 1979; Hedrick, Bickman, & Rog, 1993). Strictly speaking, one can only generalize to the accessible population from which this researcher’s sample was drawn. Several critical aspects of the populations used must be compared in order for the populations to be deemed similar. The environmental conditions also must be examined. Campbell and Stanley (1966) investigated factors that could jeopardize external validity.

Interaction effect of testing:

One factor that could jeopardize external validity is the reactive or interaction effect of testing (Campbell & Stanley, 1966). This occurs where a pretest might increase or decrease the participant’s responsiveness to the experimental variable and thus make the pretested population’s results unrepresentative of the effects of the experimental variable. This threat is considered to be minimal in this design because a pretest was not used. Therefore, it is arguable the population used may better represent the unpretested universe from which the respondents were selected.

Interaction effects of selection. According to Campbell and Stanley (1996), the interaction effects of selection refers to “the limitation of the effects of the experimental variable to that specific sample and the possibility that this reaction would be untypical of the more general universe of interest for which the naturally aggregated exposure group was a biased sample” (p. 41). It is impossible to control all the variables of selection due to realities of life (funding, participant availability, human variability, etc.). This threat warranted concern but controls were added. Although randomization or matching was not possible, and intact groups had to be used for participant selection, a larger number of participants was used (N = 100). The sample included teachers serving varied socioeconomic and geographical locations. Explicit description of the sample population and study framework was provided. The study design and instrument were cautiously fashioned. The cover letter operationalized the definitions used for the survey’s terminology, the survey was devised under specific guidelines, particular criteria were set for the participants, application and scoring of the survey was regimented, and bias of data interpretation was knowledgeably avoided. Furthermore, throughout the study, the researcher was cautious not to generalize any findings beyond the intended teacher population of the four schools selected for the study.

Experimental arrangements:

The confounding effects of the experimental arrangements might also jeopardize external validity (Campbell & Stanley, 1966). The artificiality of an experimental setting and the participants’ knowledge that they are participating in an experiment threaten representativeness and generalization. This researcher’s choice of self-administered questionnaires and repeated assurance of participant confidentiality substantially diminished this threat. This researcher was absolutely resolute not to treat any participant in a substandard fashion. All participants were provided the same materials, information, and consideration.

Multiple treatment interference:

Multiple treatment interference, or the confounding effect of pretesting (Campbell & Stanley, 1966), was controlled in this design. No pretesting was intended in this research study. The pilot test was used strictly to pilot the survey instrument and process. The results were not used in the study. Special care was taken to disallow any participant in the pilot study from retaking the survey. Any risk of the application of the interview survey in addition to the initial self-administered survey, changing the participants’ behavior–and therefore the results– were also controlled by the design. The choice to select the interview participants randomly, from the entire population being studied, greatly reduced this threat, and enhanced the validity of the study’s findings.

Statistical analysis

The data will be analyzed using excel. Descriptive statistic are used to analyze continuous and categorical data and presented in the form mean, standard deviation and percentage, while proportions are analyzed using chi-square test. To measure the reliability cronbach’s alpha will be used.