Please enable JavaScript.

Coggle requires JavaScript to display documents.

Educational Research and Statistical Significance Rick Dawson Moreland…

- - - - 2. Descriptive – Provides information about conditions or events occurring in the present. Example: A survey of school buildings to determine the state of facilities in a typical school.
        
        3. Correlational – Searching for relationships between variables by measuring statistical association. Example: Attempting to correlate teacher job satisfaction with other variables such as salary, benefits, environmental factors, etc.
        
        4. Causal – Attempts to find or suggest causal links between variables by observing present conditions and researching historical data in order to find possible causal relationships. Example: A study of drop-out rates of secondary school students using data dating back from the previous decade.
        
        5. Experimental – In which variables used to define one or more ‘causes’ are manipulated to produce observable ‘effects’ on other variables. Example: A comparative study of the effectiveness of two new textbooks by random assignment among three groups of teachers and students, with each of the first two groups being given one of the new texts, and the third acting as a control group.
        
        6. Case study – Case studies fall into two distinct categories: a) In-depth studies of specific students, classrooms, or schools, which are intended to describe how the overall culture affects education, and the interactions between students and other individuals of interest. Example: An in-depth study of friendships among students of a particular class. b) Applying quantitative research methods to non-probability samples (i.e., samples gathered in a process in which not all the individuals in the population have equal chances of being selected). Results from this type of case study may not necessarily be generalizable to wider populations. Example: A survey of the reading achievements of students in a particular location or region.
        
        7. Ethnographic – Typically describes events that occur within the life of a studied group, with a particular focus on how individuals interact within the context of social or cultural factors characteristic of the group, such as norms, beliefs, or rituals. The researcher usually is to some degree a participant in the group life in order to draw a better understanding of member interactions. Example: In order to track and detail daily interactions and tasks of a school principal, a researcher is assigned as an ‘assistant principal’ as means of participating fully in the daily life of the school.
        
        1 more item...
    - - 1. Descriptive – Often concerned with comparing existing school conditions with legislated standards, conditions in other school systems, or current conditions in several sectors of a single school system. Such questions may address the quality and condition of school buildings, supplies and equipment, teacher housing, or academic achievement in core subjects.
        
        2. Correlational – Checks for associations between variables that might indicate causation. Great care must be exercised in such studies due to the ease with which correlation can be confused with actual causation. Some examples might include comparative studies of how student groups perform academically in school environments in which facilities are in good or poor condition, have more or less equipment, etc.
        
        3. Causal – Typically considered of greatest importance to educational planner, causal questions go beyond correlational questions by seeking to determine more concrete links between isolated factors (“all other things being equal, does a variation in X produce a different educational outcome in a student group,” etc.). Such a study might compare outcomes between two groups of students using different textbooks, or measuring academic success among students of different socioeconomic status. (Postlethwaite, 2005)
      - Sequential Stages in the Research Process
        
        Stage 1: General and specific research questions – Once a general subject of research has been identified, a specific question must be crafted and finalized target as exactly as possible the target population, and the variables and other factors included in the study.
        
        Stage 2: Literature review – This is done to establish the current state of affairs within the given area of study. This review does not simply summarize previous research studies and their findings, but rather aims to identify essential issues revealed by this research. The review describes what has already been written in the subject area, how it has been received by other scholars, the major findings across studies, and the key points of debate and contention. Previous studies help identify existing knowledge gaps, theoretical points in need of empirical study, and may provide useful models in designing new research.
        
        Stage 3: Research design – Specific research questions will require different methods of examination, and a decision must be made whether to design an experimental study or a survey. In the event a survey design is chosen, a further decision must be made between longitudinal (data collection of a sample occurs at various points of time) or cross-sectional (data is collected at a single time point) surveys. When the researcher has determined which variables are to be gathered on which data, data collection units (are we studying students, teachers, principals, etc.) and methods (do we use surveys, observations, interviews, etc.) must be defined narrowly, along with all the details of data collection in the study.
        
        Stage 4: Instrumentation - In the absence of existing data from previous studies (“secondary data”), primary data must be collected. Once the questions surrounding data collection in the research design phase have been decided, the researcher can begin constructing the instruments to be used for collection. That is, if a questionnaire is to be used, then it must be written. Research instruments should then be reviewed by experienced researchers to ensure suitability in gathering the desired data.
        
        Stage 5: Pilot testing – This is where instruments and data collection and recording techniques are testing by administering them to a sample population of individuals similar to those that will be targeted in the formal study. This can help reveal weaknesses with instruments and procedures in order that they be corrected before final data collection.
        
        1 more item...
        
        Connection: Quantitative and Qualitative Data
        
        Definitions
        (Macalester College, 2020)
        
        Quantitative Data: used when a researcher is trying to quantify a problem, and can either be counted or compared on a numeric scale. For example, it could be the number of students in a given school or grade level, or summative assessment scores. This data is usually gathered using instruments like questionnaires utilizing rating scales or multiple-choice responses, or gauges such as thermometers which yield empirical measurements. Statistical analysis software is often used to analyze quantitative data.
        
        Qualitative Data: usually collected using questionnaires, interviews, or observation, in an attempt to describe qualities or characteristics and frequently appears in narrative form. Due to its subjective nature, qualitative data may be difficult or impossible to precisely measure and analyze, but it can still be highly useful in determining how subjects perceive specified issues or questions. Qualitative data might be gathered using a questionnaire with open-ended questions which allow respondents to describe their reactions in their own words.
        
        Connection: Education Research Stages 6 and 7: Data collection and analysis – Muftahu Jibirin Salihu notes that, “Often times, the approach in collecting data determines the method for analyzing the data and that serves as the link between the two. For example, if a researcher applied the qualitative research approach, then most likely, the researcher would also apply qualitative means of analyzing research data such as content analysis for data derived from in-depth interviews.” (Salihu, 2016)
- - - - 3. Select a probability of error level (Alpha level) - The researcher has to be vigilant against mistaking the relationship between the two variables.
        
        A Type I error occurs when an assumed relationship does not in fact exist. In other words, the researcher accepts the research hypothesis when the null hypothesis is true. The probability of making a Type I error is called Alpha.
        
        A Type II error is the opposite case in which the researcher embraces the null hypothesis when in fact the research hypothesis is true. The probability of this happening is called Beta.
        
        In general terms, reducing the likelihood of Type I error increases the chance of Type II, and the converse is also true. Most researchers make more effort to reduce Type I errors on the general assumption that they’re less harmful than Type II: A Type II error may result in a missed opportunity to improve a situation, but a Type I error may prompt a change for the worse.
        
        Alpha level refers to an acceptable risk level for Type I errors. In the social sciences, this is typically .05, or 5%--a 95% level of confidence, meaning that a true null hypothesis will be rejected 1 time in 20. (CSULB, 2020)
        
        4. Select and compute the test for statistical significance
        
        5. Interpret the results – The test for significance yields a “pass/fail” measurement in which the researcher may reach a conclusion as to the validity of the results. The Alpha level is the threshold value against which p-values are held. The p-value is a measure of probability too, but unlike Alpha, p-value indicates the likelihood that an observed statistic is the result of random chance. The smaller the p-value, the less likely the observed data is due to chance. To determine the statistical significance of an observed outcome, Alpha and p-value are compared:
        
        If the p-value is less than or equal to alpha, we reject the null hypothesis, and conclude that the result is statistically significant. In such a case, we are reasonably sure that the observed sample is not due to chance alone.
        
        Where the p-value is greater than alpha, we fail to reject the null hypothesis, concluding that the result is not statistically significant. In this case, we are reasonably sure that the observed data can be explained by chance alone.
        
        Implication – The smaller the value of alpha is, the more difficult it is to claim statistical significance; conversely, larger alpha values make it easier to claim that a result is statistically significant—and there is also a higher probability that what we observed can be attributed to chance. (Taylor, 2019)
        
        1 more item...