Investigación y metodología científica, más allá del TFM - Rafael Bisquerra

Characteristics of Scientific Research in Education

Overview of Today's Content

• The session will cover the characteristics of scientific research applied to education, the research process and its phases, and methodologies including quantitative and qualitative approaches.

• Emphasis is placed on providing clear ideas that are useful rather than overwhelming with information. This approach is summarized by the phrase "in a mountain set," indicating depth over breadth.

Distinction Between Knowledge Types

• A distinction is made between doxa (subjective opinion) and episteme (objective knowledge), highlighting that scientific research aims for objective understanding through the scientific method.

• Personal beliefs or opinions do not equate to scientific knowledge; true knowledge comes from systematic investigation and evidence collection.

Characteristics of Scientific Research

• Scientific research follows a structured methodology known as the scientific method, which involves empirical knowledge based on observation and experimentation. This leads to data accumulation essential for analysis.

• The goal is to achieve objectivity through specialized knowledge, which differs from general understanding, emphasizing precision and communication within the scientific community.

Understanding Probability in Science

Nature of Scientific Knowledge

• Scientific claims are probabilistic rather than absolute; they express likelihood based on variables rather than certainties, often ranging between 0% to 100%. For example, a probability of 98% indicates high confidence but not certainty.

• The concept of probability underscores that science operates within a framework of uncertainty, where findings can never be definitively proven but can be supported by evidence within certain limits.

Phases of the Research Process

Initiating Research

• The first phase in any research process involves clearly defining the problem at hand, which serves as a catalyst for further investigation—illustrated by current issues like COVID-19 prevention strategies.

Understanding the Role of Literature Review in Research

Importance of Probability in Research

• The speaker emphasizes that while conducting research, there is never a 100% guarantee of success; instead, researchers work within probabilities to address issues.

Conducting a Literature Review

• A literature review involves examining existing research on a topic. The speaker highlights the ease of accessing bibliographic resources through the internet today.

• Scientific databases and journals are crucial for finding relevant articles. Each scientific field has numerous journals that publish pertinent studies.

• Researchers often face an overwhelming amount of literature, making it challenging to read all available articles on a given topic.

Challenges in Accessing Information

• The abundance of new publications can create a "Alice in Wonderland" scenario where researchers must keep running just to stay in place with their reading.

Focusing Research Topics

• Narrowing down topics can enhance the specificity and relevance of information gathered. For example, focusing on ADHD among children in Catalonia allows for targeted insights.

Establishing a Theoretical Framework

• A solid theoretical framework is essential for guiding research. For instance, understanding emotional education's role in preventing violence can inform intervention strategies.

Strategies for Violence Prevention

Types of Violence Addressed

• The discussion includes various forms of violence beyond bullying and gender-based violence, emphasizing the need for comprehensive prevention strategies.

Emotional Regulation as Key Factor

• Impulsivity and unregulated anger are identified as significant triggers for different types of violence, highlighting the importance of emotional regulation training.

Collaborative Efforts Required

• Effective violence prevention requires collaborative efforts among educators and structured emotional education programs implemented regularly.

Formulating Hypotheses Based on Literature

Developing Hypotheses from Theory

• Researchers can formulate hypotheses based on their theoretical frameworks. For example, implementing an emotional education program may reduce violent behaviors among adolescents.

Methodology: Data Collection Techniques

Importance of Methodology

Research Process and Hypothesis Formation

Conclusions in Research

• The conclusion of research typically involves confirming or rejecting a hypothesis, but it often leads to new questions for future investigations.

• Scientific research is an ongoing process where multiple researchers contribute data that not only answers existing questions but also generates new inquiries.

Theoretical Framework

• A theoretical framework is derived from prior research and serves as the foundation for formulating hypotheses and methodologies.

• For instance, current research on COVID-19 relies on previous studies in epidemiology, while violence prevention studies are based on factors triggering violence.

Importance of Further Questions

• It is considered valuable when research conclusions lead to additional questions for further investigation, emphasizing that each study contributes to a larger body of knowledge.

• Researchers often discover aspects that remain unanswered, prompting them to suggest directions for future studies based on their findings.

Methodology and Objectives

• Discussion about expanding categories within methodology will be addressed later; objectives can be formulated during the methodology phase or hypothesis formation phase.

• Objectives stem from the theoretical framework and can lead to hypothesis development; clarity in structuring these elements is essential.

Hypothesis Formulation Debate

• The formulation of hypotheses can be contentious; some methodologies view it as crucial while others do not prioritize it as highly.

• There are varying opinions regarding whether hypothesis formulation should follow the theoretical framework or serve as an initial methodological step.

Flexibility in Research Design

• Research frameworks should not be seen as rigid constraints but rather as guidelines that allow flexibility in addressing important elements throughout the study.

Understanding Methodology and Hypothesis Rejection

Flexibility in Methodological Phases

• The speaker discusses the importance of flexibility within methodological phases, emphasizing that these phases are part of a continuous process.

• A suggestion is made to focus on key phrases that guide the methodology, highlighting the need for clarity in explaining complex concepts.

Formulating Research Problems

• Questions play a crucial role in formulating research problems; they can be articulated verbally to clarify issues.

• The formulation of a problem often begins with a simple question, which can then be elaborated upon to develop specific objectives.

Importance of Hypothesis Rejection

• Rejecting a hypothesis should not be seen as failure but rather as an opportunity to explore new questions and refine theoretical frameworks.

• Historical context is provided through an example from the 16th century regarding the geocentric vs. heliocentric theories, illustrating how hypothesis rejection led to paradigm shifts.

Paradigm Shifts and Scientific Inquiry

• The transition from geocentrism to heliocentrism exemplifies how rejecting established hypotheses can lead to revolutionary changes in understanding.

• New theories emerged as scientists questioned existing beliefs about the universe's structure, leading to further exploration and discovery.

Conceptual Mapping in Research

• A conceptual map is recommended for organizing theoretical frameworks, helping researchers visualize relationships between topics and subtopics.

• Creating such maps clarifies complex phenomena by breaking them down into essential elements relevant to the research problem.

Methods of Inquiry: Inductive vs. Deductive Logic

• The session transitions into discussing inductive and deductive methods; inductive logic relies on empirical observation while deductive logic follows syllogistic reasoning.

Inductive Reasoning and Sampling in Research

Understanding Induction in Everyday Contexts

• The act of turning on a light switch exemplifies inductive reasoning; one does not need to test every switch to conclude that they function to turn off the light.

• This principle extends to other objects, such as mobile phones, where observing one phone's functionality allows us to generalize about all similar devices.

Introduction to Inductive Statistics

• Inductive statistics involves drawing conclusions from samples rather than entire populations, which is crucial for practical research applications.

• For example, opinion polls often survey a small sample (e.g., 5,000 people) from a larger population (e.g., 30 million voters in Spain) to predict voting trends.

Generalization and Sample Representativity

• It’s important to understand that not everyone will be surveyed; thus, conclusions are based on probabilities derived from sampled data.

• Scientific research typically employs both inductive and deductive reasoning methods, allowing researchers to formulate hypotheses based on general principles.

Methodology in Quantitative Research

• Key characteristics of quantitative methodology include defining the target population and selecting an appropriate sample for study.

• An example is defining a population of children aged 10-12 with learning difficulties and then selecting a representative sample for analysis.

Challenges in Sample Selection

• Researchers often face limitations regarding resources and time when accessing subjects; hence they may only reach a fraction of the defined population.

• The power of generalization depends significantly on how representative the selected sample is concerning the broader population being studied.

Importance of Sampling Methods

• Reflecting on how to select a representative sample is critical; researchers must consider various sampling methods available.

• Different sampling techniques exist: random sampling is ideal but can be complex; accessibility sampling may be used when resources are limited.

Types of Sampling Techniques

• Stratified sampling involves dividing the population into subgroups (strata), ensuring diverse representation across different age groups or demographics.

• Intentional sampling targets specific individuals based on predetermined criteria relevant to the research question at hand.

Accessibility Sampling Considerations

• Accessibility sampling occurs when researchers cannot use random methods due to constraints but still aim for some level of representativeness within their accessible group.

Understanding Constructs and Variables in Psychopedagogy

Distinction Between Construct and Variable

• The discussion begins with the importance of distinguishing between constructs and variables, emphasizing their relevance in psychopedagogical sciences.

• A construct is defined as something that cannot be directly observed; for example, intelligence can only be inferred through behavior rather than seen directly.

• Personality is another complex construct; behaviors provide indirect insights into a person's personality traits.

Measurement Instruments

• Intelligence and personality are measured using specific tests (e.g., intelligence tests or personality assessments), which serve as variables within the broader constructs.

• Different instruments can yield varying data about the same construct, highlighting the need for careful selection of measurement tools.

Types of Variables

• The distinction between dependent and independent variables is crucial. Independent variables influence outcomes, while dependent variables are affected by these changes.

• An example illustrates this: applying an emotional education program (independent variable) aims to reduce violent behavior (dependent variable).

Data Collection Methods

• Various methods exist for measuring dependent variables in psychopedagogy, including tests, questionnaires, interviews, and observational techniques.

• Reliable data collection instruments are essential to observe changes effectively over time.

Clarifications on Measurement Techniques

• A moment is provided for questions regarding constructs, variables, and data collection methods to ensure understanding before progressing further.

Understanding Constructs and Variables in Research

Theoretical Frameworks and Definitions

• Constructs are defined within a theoretical framework, which shapes their conceptual definitions. This highlights the importance of establishing a solid theoretical foundation for research.

• Intelligence is operationally defined through specific tests, such as the Weber test, illustrating how different definitions can exist based on context and application.

Variables in Research Methodology

• Dependent variables are linked to intervention programs, while independent variables may include techniques like interviews. The classification depends on their role in the research methodology.

• Interviews can serve dual purposes: as an intervention technique or as a means to assess behavior. Their function varies based on the research design.

Measurement Techniques

• Independent variables in therapy are often described through protocols that outline diagnostic steps and procedures, emphasizing structured approaches to interventions.

• Implementing pre-tests and post-tests is recommended for evaluating interventions. Pre-tests can encompass various assessment tools beyond just one instrument.

Inclusion and Exclusion Criteria

• Inclusion criteria define individuals who meet specific characteristics necessary for participation in a study, while exclusion criteria eliminate those who do not meet these requirements.

• Clear definitions of inclusion and exclusion criteria are crucial to avoid redundancy in sample selection, ensuring that only relevant subjects are included.

Experimental vs. Quasi-experimental Designs

• Experimental designs involve random sample selection for rigorous methodological standards, whereas quasi-experimental designs may lack this randomness but still aim for valid results.

Understanding Experimental Design

Accessibility and Representativity in Samples

• Accessing samples based on accessibility criteria often raises concerns about their representativity. Even in experimental designs, the representativity of samples is frequently debatable.

Pre-Test and Post-Test Design

• The pre-test/post-test design involves measuring a variable before (pre-test) and after (post-test) an intervention or treatment to assess changes.

Experimental vs. Control Groups

• In a three-group design, there are experimental and control groups. The experimental group receives the independent variable (e.g., therapy), while the control group does not.

Measurement of Variables

• The dependent variable is measured at both pre-test and post-test stages. For example, measurements may be denoted as x1 for the experimental group's pre-test data.

Data Collection Techniques

• Various data collection methods can be employed, such as tests, interviews, or observations. These techniques help gather relevant information from educational settings.

Application of Treatment

• The independent variable (e.g., emotional education program) is applied to the experimental group over a specified duration, with outcomes assessed through post-tests.

Mortality in Experimental Designs

• Differences in subject numbers between pre-tests and post-tests indicate "mortality" in experiments; this term refers to participants not being present rather than actual death.

Importance of Analyzing Mortality Rates

• High mortality rates can significantly impact research validity. Understanding reasons behind participant dropout is crucial for accurate conclusions.

Repeated Measures Design Clarification

• A repeated measures design can involve two or more measurements over time; it’s essential to clarify that a pre-post test design is a specific case within this broader category.

Significance of Sample Characteristics

Experimental Design and Statistical Analysis

Understanding Experimental Groups

• The educational center can serve as an experimental group, ensuring that 330 distinct centers share the same characteristics. This is crucial for comparing groups where one receives the independent variable while the other does not.

• The terms x1 and x2 represent pre-test and post-test measurements of the experimental group, respectively. These symbols help clarify discussions about data comparisons between groups.

• X3 and x4 denote pre-test and post-test measurements of the control group. The goal is to compare these with the experimental group's results to assess effects accurately.

Key Concepts in Measurement

• It’s essential that x1 (experimental pre-test) and x3 (control pre-test) are equivalent to ensure valid comparisons. Differences between x1 and x2 indicate effects from the independent variable.

• The absence of effect in the control group (x3, x4) should demonstrate no changes when not exposed to the independent variable.

Data Analysis Considerations

• Only subjects who completed both tests should be included in analyses; this avoids issues related to experimental mortality affecting results.

• Pre-test data (x1) should show no statistically significant differences among subjects, indicating a reliable baseline before treatment application.

Types of Tests Used

• Examples of tests for pre-tests and post-tests include intelligence tests, academic performance assessments, or emotional competency evaluations. Consistency in testing instruments is vital unless prior exposure affects outcomes.

• In cases where prior knowledge may influence results (e.g., history questions), different but comparable instruments might be necessary for accurate assessment.

Clarifying Experimental Design

• An example of an experimental design involves selecting a classroom for research access rather than random selection across all classrooms, which could introduce bias.

• A clear distinction exists between non-randomly selected groups versus randomly assigned ones within a population framework.

Introduction to Statistical Concepts

• Transitioning into statistical analysis requires foundational knowledge; those unfamiliar with statistics may find it challenging to follow advanced concepts discussed later on.

• Emphasis on understanding contrasts between experimental and control groups during both pre-tests (x1 vs. x3) and post-tests is critical for hypothesis validation.

Statistical Significance in Experimental Design

Understanding Statistical Differences

• The third contrast examines statistically significant differences between the experimental group’s pre-test (x1) and post-test (x2), as well as comparing the control group with other supports.

• The fourth contrast may show non-significant differences, analyzed through various statistical techniques, particularly ANOVA for sample designs.

Group Dynamics in Research

• In educational research, both experimental and control groups can consist of multiple classes; it's essential to identify these groups accurately.

• A second control group that does not undergo a pre-test is introduced to mitigate potential biases affecting post-test results.

Data Management and Analysis

• Initial steps include creating a data matrix followed by descriptive statistics and hypothesis testing to identify differences among groups.

• It is recommended that tests be administered via computer to minimize transcription errors when inputting responses into a data matrix.

Error Detection in Data Entry

• Common transcription errors include incorrect age entries or misrecorded response options, which should be minimized and detectable during analysis.

Structuring the Data Matrix

• The data matrix can take forms like Excel spreadsheets or statistical software formats (e.g., SPSS), where each row represents an individual subject and columns represent variables.

• Essential identification variables include subject ID, sex, location, profession, marital status, etc., tailored to the study's sample characteristics.

Completing the Data Matrix

• After entering dependent variable data from pre-tests into designated columns, ensure all relevant information is organized systematically for analysis.

Visualizing Data for Analysis

• When structuring the complete data matrix, place experimental group rows first followed by control group rows for clarity in analysis processes.

• While demonstrating a data matrix might be challenging without visual aids, envisioning it as an Excel sheet helps conceptualize how subjects' responses are organized.

Utilizing Statistical Software

Statistical Analysis of Age and Test Results

Age Distribution and Statistical Measures

• The age distribution shows a majority at 18 years, with the mean age calculated as 17.43 based on a sample size (n) of 60.

• The median age is also noted as 18, while the mode reflects that 18 is the most frequently occurring age in the dataset.

• An error in data entry was identified where a subject's age was incorrectly recorded as 61 instead of 16, highlighting the importance of data accuracy.

• Following descriptive statistics, hypothesis testing will be conducted using t-tests to analyze pre-test results.

Pre-Test Results Comparison

• The experimental group had a mean score of 732, while the control group scored an average of 684; however, these differences were not statistically significant.

• Detailed analysis indicated that for equality of means, t = 199 was observed.

• A bilateral significance value of 0.52 suggests no statistically significant differences between groups since it exceeds the threshold of 0.05.

Understanding Statistical Significance

• If significance values exceed 0.05, differences are interpreted as not statistically significant; values below indicate significance.

• The probability of error when rejecting null hypotheses is typically set at alpha = 0.05; thus, a significance level above this indicates acceptance of null hypotheses.

Post-Test Analysis and Findings

• In comparing post-test scores between groups (experimental: mean = 732 vs control: mean = 684), statistical tests showed no significant difference initially but later revealed slight effects from interventions.

• After applying statistical tests post-intervention, results indicated that differences became statistically significant with p-values approaching zero (e.g., p < .001).

Conclusion on Intervention Effects

• A very low probability error (noted as .000 or similar formats with three or four zeros after decimal points), indicates strong evidence against null hypotheses suggesting meaningful intervention effects exist between groups.

• For pre-test comparisons within control groups showing means close together (684 vs. 676), no significant differences were found with high probability errors around .727 indicating lack of effect from interventions prior to treatment application.

Analysis of Experimental Group Effectiveness

Statistical Significance in Experimental Results

• The experimental group post-test shows a significantly higher mean compared to the other three groups, indicating the treatment's effectiveness.

• In pre-test comparisons, no significant differences were found between the experimental and control groups, suggesting they were initially equivalent.

• Post-test results reveal significant differences between the experimental and control groups, confirming that the intervention had an impact on the experimental group.

• A discussion is opened for clarifications regarding these findings and their implications for statistical analysis.

Understanding Sample Size and Its Impact

• A question arises about why the sample size (n) in post-tests is larger than in pre-tests; this may be due to participant attendance variations.

• Small sample sizes often do not yield significant differences; thus, larger samples are preferred for reliable results.

• Clarification is sought on conducting tests only when there are no significant pre-test differences between groups.

Application of Statistical Tests

• Different statistical tests can be applied depending on data characteristics; both parametric and non-parametric tests can yield similar results under certain conditions.

• When sample sizes are small (less than 30), non-parametric tests are recommended to ensure validity of results.

• Discrepancies in outcomes from different statistical methods may indicate an artifact effect rather than true data relationships.

Validity Concerns in Research Design

• A student raises concerns about reliability and validity within quantitative research designs, particularly regarding maturation effects as potential sources of invalidation.

• Maturation processes must be controlled through careful timing of interventions to avoid skewed results due to natural development over time.

Addressing Maturation Effects

• The influence of maturation on pre-test and post-test differences is acknowledged as a possibility that could confound intervention effects.

Understanding Experimental Design and Analysis

Overview of Experimental Control

• Discussion on the importance of maturation in experimental design, emphasizing the need for control groups without pre-tests as a method to manage variables.

• Introduction to complex designs that incorporate multiple covariates and statistical techniques such as multivariate analysis of variance (MANOVA).

• Highlighting the necessity to control all potential variables affecting results, particularly maturation, when selecting appropriate experimental designs.

Types of Studies

• Differentiation between descriptive studies aimed at describing phenomena and developmental studies like longitudinal and cross-sectional studies.

• Explanation of longitudinal studies tracking subjects over time versus cross-sectional studies analyzing large samples at a single point in time.

• Description of cohort studies that evaluate changes within a specific group over an extended period, noting participant attrition.

Analytical Techniques

• Overview of comparative causal analyses that allow for group comparisons to identify cause-effect relationships among observed effects.

• Introduction to correlation techniques including parametric methods like Pearson's correlation and non-parametric alternatives such as Spearman's rank correlation.

Survey Research Methodology

• Emphasis on defining objectives for surveys, determining necessary information, and selecting target populations for data collection.

• Importance of designing questionnaires through pilot testing with small groups to ensure clarity and effectiveness before full deployment.

Response Options in Surveys

• Discussion on various response formats including simple categorical responses and multiple-choice options tailored to gather nuanced data from participants.

• The significance of accurately establishing response categories based on survey topics while ensuring they are exhaustive and mutually exclusive.

Data Analysis Techniques

• Review of statistical software applications used for data analysis, highlighting outputs from programs like SPSS for managing datasets effectively.

• Introduction to evaluative research focused on program assessment through needs analysis, implementation evaluation, and outcome measurement using models like pre-test/post-test with control groups.

Program Evaluation Models

• Presentation of the ZIP model (Context, Input, Process), which analyzes program components essential for improving educational interventions.

Understanding Program Evaluation

Importance of Evaluating Processes and Context

• Emphasizes the need to evaluate not just the product of a program but also the process involved, highlighting that understanding the context is crucial before implementation.

Research Methodology in Program Evaluation

• Suggests that individuals should identify a specific problem for investigation and seek relevant literature to frame their theoretical approach.

Designing Research Methodology

• Discusses choosing between quantitative and qualitative methodologies, noting that qualitative methods focus more on observation, interviews, and group discussions.

Structuring Research Work

• Outlines essential phases of research work: introduction, problem statement, theoretical framework, research design, data collection process, data analysis, and conclusions.

Bibliography and Literature Review Guidance

• Responding to a request for bibliographic resources related to educational methodology; suggests looking for books with titles focused on research methodology across various fields.

Recommendations for Finding Resources

• Advises searching libraries or online platforms for extensive literature on research methods applicable in education and social sciences.

Sending Bibliographic Lists

• Offers to send a bibliographic list via email upon request from participants interested in further reading materials.

Conducting Literature Reviews

• Addresses inquiries about conducting literature reviews; recommends following established norms and offers assistance with guidelines on how to perform effective reviews.

Key Steps in Literature Review Process

• Highlights the importance of knowing one's field's documentation sources and consulting experts when determining appropriate bibliographic searches.

Keywords for Effective Searches

• Encourages using 3 to 5 keywords relevant to the topic when conducting bibliographic searches to enhance search efficiency.

Example of Keyword Usage

Research Methodology in Emotional Competence

Combining Keywords for Research

• The speaker discusses the importance of combining keywords such as "emotional competencies," "anger," and "violence" to enhance search results in academic databases.

• Emphasizes using Google Scholar for bibliographic searches, highlighting its accessibility and wealth of scientific literature.

Interview Techniques in Addiction Studies

• A question arises regarding the number of trust-building questions needed during interviews with addiction patients; there are no strict norms, as it varies by interview type.

• The speaker suggests preparing between 10 to 40 questions, including control questions that rephrase inquiries to assess response consistency.

Statistical Considerations in Small Samples

• Discussion on whether experimental and control group designs apply to case studies or small samples; a minimum number of subjects is necessary for statistical validity.

• The law of large numbers indicates that statistical methods work best with groups larger than 30; smaller groups may require non-parametric tests.

Validity of Single Subject Research

• It is possible to conduct serious scientific research with just one subject (n = 1), citing Piaget's research methodology as an example.

• Clarifies misconceptions about minimum sample sizes, stating that valid research can be conducted even with a single participant.

Conclusion and Further Learning Resources

• The speaker expresses gratitude for participation and encourages further reading on research production through articles and books for deeper understanding.