Saturday, December 31, 2011

Classification of research designs

Quantitative researchers have developed different types of research designs to facilitate their studies. Their research procedures include organization of the variables, selection of samples, schedule for data collection and techniques for statistical analysis.

Similarities between research designs have allowed to to classify them in various types. The criteria of the use of experiment have allowed to classify the designs in experimental and non-experimental research.  In non- experimental design the researcher studies existing phenomena without intervention in the structure of these phenomena. On the other hand the experimental design require some kind of experiment with the intervention of the researcher. The various types of non-experimental design are: descriptive, causal-comparative and correlational.

Research designs are also classified according to their purpose. Educational studies are undertaken according to four different purposes: description, prediction, improvement and explanation.

If the purpose is description, two types of research design can be used: descriptive and longitudinal. Descriptive research is used when phenomena are studied at one point in time. Longitudinal design is used to study changes that occur in phenomena over time.

If the purpose is prediction, correlational design is used.

When the purpose is explanation a causal-comparative design is used. The design involves finding cause and effect relationships between variables. Correlation and experimental design are also used. I would also add that descriptive design can be used because when you describe something you also explain it.

Finally if the purpose of the design is improvement, experimental design is used.  The reason why experimental design is used is because designing an intervention and following its effects is a kind of experiment.

Saturday, December 24, 2011

Statistical techniques

Definition of Statistics

Statistics are mathematical techniques  for analyzing numerical data to accomplish various purposes. For example the calculation of a mean leads to a single score that represents many scores such as the scores of all students who took a particular test.

Types of Statistics

There are four types of Statistics: descriptive, correlational, inferential and psychometric.

Descriptive Statistics

Descriptive Satistics are mathematical techniques for organizing and summarizing a set of numerical data. In Educational Research they are used to describe educational phenomena. They describe the score on a single variable

Correlational Statistics

Correlational Statistics are used to describe relationships between two or more variables.

Inferential Statistics

Inferential Statistics are mathematical techniques for using probabilities and information about a sample to draw conclusions about the population from which the sample originated.

Psychometric Statistics

They are Statistics used to describe the psychometric properties of tests and he appropriateness of information and uses of outcomes from tests and other measures.

Example of Statistical Analysis in a Research Study

Statistics are used in virtually all quantitative studies but in many qualitative studies also. An example is the case study of reading groups in a first grade classroom conducted by James Collins. The study was part of a large ethnographic study of language differences between working class black children and middle class white children in their home and school environments.

Types of Scores 

Measurements in educational research are expressed usually in three forms: continuous scores, ranks, or categories. It is important to understand the differences between the score types because the forms in which the scores are expressed reflect the choice of statistical analysis procedure. If the research data consist of continuous scores on two groups group differences are analyzed by calculating a mean score and a statistic known as t. However if the scores are in thew forms of categories the group differences are analyzed by a Chi-square test, which compares categories frequencies between two or more groups. There are 3 types of scores: continuous, age and grade equivalents and rank.

Saturday, December 17, 2011

E-learning in Africa and China

Submitted by Michael Trucano on Fri, 12/16/2011 - 11:26

Earlier this year, over 1700 participants from over 90 countries attended eLearning Africa (previous blog post here) to share lessons and make contacts at what has evolved into perhaps the continent's premier annual knowledge sharing event related to the use of ICTs in education. Not surprisingly, Tanzania led the way in terms of attendance by its nationals, followed by its East African neighbors, with South Africa and Nigeria not too far behind.

One nationality was largely noticeable through its absence: the Chinese.  Why do I mention this? Outside the conference, signs of growing cooperation between Tanzania and China (and India, whose Prime Minister was in Dar the same week on a state visit) were hard to miss, and indeed, the increasing 'presence' of China across Africa is undeniable, and the topic of much reporting, scholarly interest and discussion, including at the World Bank. Looking around the conference itself, this cooperation wasn't immediately in evidence related to international cooperation around the use of educational technologies.  Participating in and listening to many conversations at the event, however, one got a bit of a different story related to potential cooperation going forward between China and a number of African countries on ICT/education issues.

While comparatively few representatives from Chinese firms and organizations participated at eLA, after engaging in a few dozen informal discussions with many MOE staff, vendors and consultants, it is clear that Chinese support for the purchase of ICT infrastructure for schools will most likely increase greatly in the coming years.  Scattered existing examples of small cooperation were cited by many people as a harbinger of things to come.  Almost every ministry of education official with whom I spoke mentioned that they had contact of some sort with Chinese officials or partners around the use of computers in schools, and expected this to increase in the near term (many remarked on how this contrasted with their dialogue, or lack thereof, with most 'traditional' donors on this topic).

Why is this potentially important? The potential for 'South-South' knowledge exchange, something increasingly championed at the World Bank, is pretty clear. At a speech last year in China talking about China's achievements with Special Economic Zones and infrastructure development, the World Bank president noted that "African countries want to learn from such success, and China is ready to help." He continued: "China’s experience can be instructive for African countries.  It also suffered from infrastructure deficits at the beginning of its development process but succeeded in putting in place world-class infrastructure -- covering both urban and rural areas.  Africa may also draw from China’s attention to rural infrastructure as a way to improving productivity and overcoming poverty."

Discussions about 'Africa' often founder, given the (obviously) tremendous diversity in situations and circumstances across the continent.  The same can be said for discussions about 'China', given its large size and great diversity. While the results from Shanghai in the latest PISA round are the envy of much of the rest of the world, the relevance of mass school computerization efforts in rural Western China may well offer insights to some African policymakers that they might not get when talking with consultants drawing on the experience of ICT use in schools in, say, Toronto or Lyon or Manchester.

Despite what appears to be growing interest in cooperation between a number of African countries and Chinese partners on issues related to putting ICT infrastructure in schools, my anecdotal impression is that lessons from Chinese experiences in using technology in education are not well known outside of China. When I mention to ministries of education around the world that I spent a few years working on an ICT/education project in China near the start of the last decade, I am almost immediately bombarded with lots of questions.

One can postulate a number of reasons for this lack of knowledge about Chinese experiences with educational technologies, including the fact that things in China are simply happening so quickly, and as a result people have been too busy 'doing' to take the time to reflect and study this experience at great length. Of course, the same could be true of most other areas of development in China, but in some ways the educational technology field seems a bit anomalous in this regard, given the intense interest of academics and policymakers in learning from Chinese experience in so many other areas. Language is also no doubt an issue here, as recent Chinese experience with educational technologies is not well documented in English and other major international languages (and if anything, seems to me to have become comparatively less so in recent years).

Through outreach activities of groups like KERIS, and in part due to a variety of cooperation efforts between the Republic of Korea and the World Bank exploring a variety of ICT/education issues, the Korean experience is slowly becoming better known to policymakers throughout East Asia, and further afield in places like Colombia, Costa Rica and Uruguay as well.

Here's hoping that the Chinese experience will become better known as well.

Epistemological issues in Educational Research

Educational Research is not an unified enterprise. Two main approaches that are studied in Educational Research are known as Quantitative Research and Qualitative Research. Quantitative Research studies samples and populations and uses statistical analysis to represent data and model the relationships between them. Qualitative Research makes little use of statistical analysis and relies on verbal data and subjective analysis. The reason for these different approaches originates from the different epistemological issues underlying scientific inquiry.

Let's see what has preoccupied epistemologists for a certain period of time. First let's define the word epistemology. Epistemology is the branch of Philosophy that studies the nature of knowledge and how it is acquired and validated. The epistemologists interested in the natural and social sciences, also called philosophers of Science, have looked for answers to some very pertinent questions. These questions are so formulated: are the objects (neutrons, self-concept) studied by researchers real? How does research knowledge different from other forms of knowledge and does it have any authority? What is theory and how can it be validated? What does it signify to find laws enabling to predict individual and group behavior? Is inquiry in the social sciences different from the inquiry in the natural sciences?

Following their long investigations about these questions during many centuries they have established different schools of thought (empiricism, phenomenology, positivism, etc). Researchers have been influenced by these streams of thinking and have come up with their own epistemological views of how research should be conducted in different branches of social sciences (Psychology, Sociology, Anthropology, etc). Educational researchers have their own epistemological approaches at this present time. Other Educational researchers have quite a different stand about these current issues and how they conduct their own research.

The model of Social Science Inquiry include the following elements:
A is the person being studied. This person's environment includes: a physical reality B and a social  reality C.
D is the researcher
E is the research report
F is the reader of the report.
This model can be illustrated by a diagram that can't be drawn here. In this diagram A, B, C, D, E and F are surrounded by ellipses. The diagram can be drawn in the following order:
On the right A is surrounded by an ellipse.
Above A on the left is the person's environment surrounded by an ellipse. Below this ellipse are B and C surrounded by an ellipse each: one on the left and the other on the right. Arrows relate the person's environment ellipse to ellipses B and C. This mini-diagram by itself shows that the person's environment is related to a physical reality B and a social reality C. In the first part of the diagram A, B and C are aligned.
Below between A and B is D
Below C  is E.
Below E is F so that D, E and F are aligned vertically. It is not important if someone cannot draw the diagram. The model is explained below and can be understood without the presence of a diagram.

One of the elements of the model in the diagram is the individual person designated by (A). Let's called her a teacher. The teacher interacts in an environment that is both physical (B) and social (C). The teacher uses a textbook which is a physical object made of ink and paper whose reality is defined by some chemical properties, The teacher uses this textbook to perform a social function which is to instruct students. Teacher's and students' roles are defined by society and constitute a social reality.

Educational researchers (D) and other social scientists such as Psychologists, Sociologists, Anthropologists, etc conduct investigations about  persons or group of persons, their environment or the interaction between people or group of people and their environment. For example Piaget study how children interact with their environment during different stages of their development. Some educational researchers study different social interactions in the classroom. Social scientists do not generally study physical reality although some might do. For example investigation has been conducted to study the relationship between the brain functions and the cognitive processes (attention, problem solving) while individuals work on intellectual tasks.

Following a study a research writes a report (E) about his findings, which is read by other individuals (F). These individuals can be other researchers, educational practitioners, policymakers, funders of the research. Different reports can be written for different audiences.

Saturday, December 10, 2011

Open Education and challenges in higher education

 The growing interest in Khan Academy, MOOCs and Stanford University’s online courses has made many in higher education realize that clear divides don’t exist any longer. The boundaries are blurring between real and virtual spaces, formal and informal learning, teachers and learners
Learning is changing, but what of education? A couple of blog posts this  week questioning the value of going to university at all are probably just the first of many.
A number of educators have been discussing these issues, as practitioners:  the opportunities and challenges of open online education, the role of the university, and our role as educators. Following is an edited draft of some of these discussions.
The growth of open online learning over the past decade has been steady. Open content, often discussed in terms of OERs (Open Educational Resources), is defined as “materials used to support education that may be freely accessed, reused, modified and shared by anyone”. The key to OERs is that they are openly licensed and thus available for use by all. The argument for using OERs is clear: if every university teaches introduction to programming, for example, then why should we all develop materials to teach this? Why not use openly available, openly licensed, excellent material, and spend more of our time on activities such as engaging with students, developing improved assessment strategies, etc.
There are many excellent sources of OERs (Open Educational Resources), including the NDLR; MERLOT; MIT OCW; OU Learning Space; OER Commons; Khan Academy; Stanford University’s online courses and more.
In terms of open online learning, MIT OpenCourseWare, Khan Academy and other video-based resources can be characterized as 1st generation, while the recent initiative by Stanford University, among others, can be considered 2nd generation, in that it includes not only learning materials, but instructional design, a learning structure and assessment – providing an experience closer to that provided within formal education. Stephen Downes recently suggested that the next generation will be widespread use of OERs along with automated, analytics-based, competency-based testing mechanisms, or open assessment. Indeed, this is precisely what OER university (OERu), among others, is setting out to do. Other open initiatives such as MOOCs and Open Badges have further potential to disrupt traditional higher education. Over 2000 people are currently participating in the #change11 MOOC “Change: Education, Learning and Technology”. Mozilla’s Open Badges project, particularly the DML competition on Badges for Lifelong Learning, is currently gaining a huge amount of attention as well.
Our challenges as educators in the further and higher education sectors? Here are just a few:
Open resources – Most students are aware of open educational resources, and these are shared widely, e.g. Khan Academy, YouTube, MIT OCW, and the recent Stanford University online courses. As educators, what are we doing to create or link to relevant online resources for students? Creating screencasts, video lectures, audio or video podcasts (and making these openly available) or linking to OERs (and OER repositories) can supplement lectures and provide students with valuable material for study and revision. Just as we refer students to the best textbooks, journals and databases, we should link to excellent, relevant, online open educational resources. Our challenge here is to create and share material in new ways, learn to use different tools, and stay abreast of online learning developments.
Open, participatory and social media – Students use social media and social networks in many ways, not least to support their studies, e.g. DropBox, Google Docs, Facebook, Twitter. Once again, as academic staff, we must look to our own practice. Are we making use of tools such as social bookmarking, social networking, web-based applications, and online curation tools to model good academic practice and to share resources with students, and with one another? Not all student work must be submitted directly and privately to the lecturer – opportunities for openness, sharing and collaboration should be considered.  We are challenged to consider using open, social tools (at least sometimes) – instead of closed, 1:1 tools – in order to open up the learning process and make it more authentic.
Emerging technologies – In the 2011 Horizon Report, mobile devices and e-books are the most current of the emerging technologies identified. How are we addressing these trends? The Horizon Report lists examples of education institutions innovating in these areas for teaching, learning and research. Even if we are not at the front of the innovation curve, we must address these emerging technologies in our programmes in a coordinated way, and communicate to our students and others how we are doing that. For example, how are we making use of mobile apps, or making our own learning content available on mobile devices? How are we facilitating students in using open access or e-textbooks?
Openness – In most undergraduate and postgraduate programmes, students are encouraged to examine their digital footprint and digital identity, and to consider the value of building a deliberate, positive, digital identity. This is a core element of digital literacies. Our students are visible to us online, and we are visible to them. As academic staff, are we open and positively visible online, as professionals? Are we modelling academic values in virtual spaces? The best way to share and publicise open educational resources is through the use of social media and social networks, e.g. Facebook, Twitter, Google+, blogs. In order to communicate and share our work and our values, our challenge is to consider our approach to openness – as individuals, as departments, and as universities.

Application of Resarch to Educational Practice

Based on some observations Educational Practice cannot rely blindly on research. Both fields have their own goals. According to D.C Philips Research uses statement about what is and Educational Practice uses statements involving "ought to be". For example Research uses statement such as : "X is Y"; "the probability for X to have the feature Y is p". Practice uses statements such as: "Person A ought to do Z to person B". It is logical  that from statements involving the use of "is" conclusions about "ought" or "should" cannot be deduced.

Questions involving "is" can be well answered by educational research while those involving "ought to" imply the use of dialogue to solve them. Researchers cannot expect their findings about "is" to be transformed in immediate change without being criticized. Likewise practitioners cannot look to research for prescriptive advice. However practitioners can use the researcher's findings in their dialogue about solutions to practical problems.

Limitations of Research knowledge

Research findings has several limitations. One of them is that results generated from a sample cannot be generalized to all the elements of a population. Some research studies do a few cases and generalization has to be done by considering each of the other additional cases. Therefore practitioners can look to research for advice but they should ask themselves: "Are these findings applicable to my situation"?

Another limitation of research knowledge is that its discoveries are filtered with a certain worldview. Studies about intervention in the classroom may have have high performance test achievement as learning outcomes while neglecting other outcomes such as self-reliance, humanitarian attitudes,etc. Some research studies are done with a certain view of teachers as proved by Lampert's observations. She stated that teachers are considered like "technical production manager" whose role is is to monitor the efficiency of learning. The teacher's role is to apply researcher's knowledge and policies without the consideration of other instructional factors.

Lampert advances a different view of teachers as dilemma managers. This view originated from her own research studies revealing that classroom teaching involves many problematic situations with competing interests that the teacher has to deal with.

Lampert's view of teaching agrees with other professionals who studied professional practice. Donald Schon is one of the most influential of these individuals. His theory stated that a "flawed model of technical rationality" dominates thinking about the relationship between research and practice. He describes the model of Technical Rationality as professional activity consisting in "instrumental problem solving made rigorous by the application of scientific theory and technique".

Shon explained the reason why this model is flawed. He stated that research in the positivist tradition deals with a "stable, consistent reality about which generalizations can be made and applied, whereas professional practice involves "complexity, uncertainty, instability, uniqueness and value conflict."

Schon and many others advise that practitioners have to engage in reflection-in-action, not in the application of research knowledge in order to deal with the "messiness" of their work. One of the chief elements of reflection-in-action "is a kind of experimentation based on the practitioner's analysis of each unique situation they confront."

Schon's model of reflection-in-action doesn't prevent the application of research knowledge for professional action. The implication of the model is that research knowledge should not be used exclusively as a basis for professional action. In fact researchers found that "the classroom is marked more by sameness of practice than by diversity and uniqueness". In fact research knowledge might allow practitioners to be in a better position to accommodate the differences among the constituencies."

 The Importance of Basic Research

Some practitioners believe that educational research is too theoretical and too focused on basic processes of learning. They think that priority should be given to applied research based on problems confronted by practitioners. This argument raises questions about the relative value of basic and applied research in education.

While the contribution of applied research to the improvement of educational practice seems obvious an important study in the field of education gives reason for reconsideration of this viewpoint. The remarkable findings of this study are related to the fact that a high percentage of basic research studies was essential to the development of current treatment of cardiovascular and pulmonary disease.


Monday, December 5, 2011

Four types of knowledge that research contributes to education


Many research studies are based on the description of natural or social phenomena by studying their form, structure, activity, change over time, relationship to other phenomena, etc. These descriptions have resulted in important discoveries. For example the observation of different parts of the universe by astronomers have resulted in the discoveries of galaxies and the structure of the universe. These discoveries have subsequently lead to the origin of the universe and its course.

The descriptive function of research relies strongly on instruments for observation and measurement. Researchers spend a great amount of time to develop instruments. Once developed these instruments are used to describe phenomena studied by researchers.

Descriptive studies increase the knowledge of education in schools. Some important books about education are based on descriptive studies for example Life in Classroom by Philip Jackson, The Good High School by Sara Lawrence Lightfoot, etc.

Some descriptive educational studies produce statistical information of interest to policy makers and educators. The National Center for Education Statistics publish descriptive studies in an annual journal called the Digest of Educational Statistics. These published studies about the delivery of education are information available to anyone concerned about the quality of education in schools.


Another type of research knowledge is the ability to predict a phenomenon that will happen at time Y from information available at a time X. For example lunar eclipses can be predicted with precision from knowledge about the relative motion of the Moon, Earth and Sun. The next stage of an embryo's development can also be predicted from knowledge about the current stage of the embryo. A student's performance in school can be predicted with a fair amount of precision by an aptitude test administered previously a year or two,

Educational researchers have undertaken a lot of prediction studies for the acquisition of knowledge about factors predicting student's success in school and in the world of work. One of the reasons of doing such research is to provide guidance for the selection of students who will be successful in some particular academic disciplines. For example Universities use the Scholastic Aptitude Test along with other data for the selection of students who are likely to be successful in their academic programs. More knowledge is needed about the level of accuracy of these tests for different groups of students to determine if new instruments are needed to improve the predictability of success in some particular fields.

Another purpose of prediction research is the identification of students who can be unsuccessful in the course of their schooling so that academic prevention programs can be put in place. For example this type of research can be used to solve the problem of school dropouts. The collection of information about students from sixth grade until graduation can provide information about the best predictions. The predictive knowledge can be used to determine sixth graders who are likely to become high school dropouts. This knowledge can be used to develop programs that can help them to be successful in schools.

Educational research has produced a large body of predictive knowledge about factors predicting issues of social importance (examples: academic success, career success, criminal conduct). Several procedures have been developed for doing predictive research.


The third type of research knowledge deals with the effectiveness of interventions. Some examples of interventions are: drug therapies in medicine, construction materials in engineering, marketing strategies in business, and instructional programs in education. Many educational research studies are realized to identify interventions or factors able to be transformed in interventions in order to improve student's academic performance. Walberg and his colleagues have summarized the results of nearly 3,000 studies on interventions or potential interventions undertaken for the purpose of improving student's performance on various measures of academic achievement. Such intervention variables are: reinforcement, reading training, cooperative learning, personalized instruction, tutoring, individualized science, individualized mathematics, etc.

Walberg's synthesis of research have demonstrated that educational researchers have found many effective interventions that can improve student's learning. However studies need to be done to improve the effectiveness of these interventions. Research has to be undertaken also to turn potential interventions into actual interventions. For example class morale is not an intervention per se. However when it is used it becomes an intervention to improve student's performance. Various research approaches are used to generate "improvement" research knowledge such as evaluation research, experimental research and action research.

Another approach for the improvement of education through inquiry has become popular in recent years. Cultural studies, a branch of critical theory, is a type of social science inquiry that investigates the power relationships in different members of a society in order to help them to deliberate from different forms of oppression. Researchers engaged in this type of inquiry should state their purpose not as improvement of education but as emancipation of some oppressed members of the educational system. A branch of historical research called revisionist theory also examines oppressive power relationships. These power relationships reflect some strong cultural and social forces that affect student learning.


The fourth type of research knowledge, explanation, is the most important one because it includes the three. Being able to explain a phenomenon researchers can describe it, predict its consequences and intervene to decrease or eliminate harmful consequences.

Researchers generally called theories the explanations about the phenomena being investigated. A theory is an explanation of a certain set of observed phenomena in terms of a system of constructs and laws that relate these constructs to each other. In other words a theory is a system that consists of a set of constructs and their relation to each other. For example, Piaget's explanation about intellectual development is a theory. Let's demonstrate it.

What phenomena does Piaget have to observe and explain? He has to observe and explain the behavior of infants and children with respect to their environment. For example Piaget observed how children of different ages responded to a particular task. The children's responses constitute a set of phenomena that Piaget has to explain by establishing a theory.

What are Piaget's theoretical constructs? First let's define a theoretical construct and its two types. A theoretical construct is a concept that is inferred from observed phenomena. It can be defined constitutively or operationally. A constitutively defined construct is one that is defined by referring to other constructs. For example Piaget's construct of conservation can be defined as the ability of an object to have some of its properties remain unchanged while other properties of the object (e.g., substance, length, volume) undergo a transformation. The notion of conservation is being defined here by referring to other constructs (e.g.,  property, transformation, or length).

An operationally defined construct is one that is defined by specifying the activities used to measure or manipulate it. For example the concept of conservation (constitutively defined above) can be defined operationally by referring to a particular task, for example, putting a constant amount of liquid into different-sized containers and then asking a child whether the amount of liquid remains the same.

Some researchers used the term variable in their investigation rather than construct. A variable is a quantitative expression of a construct. Variables are usually measured in terms of scores on an instrument of measure such as an achievement test or an attitude scale or in terms of categories of construct (e.g., public vs. private schools).

Let's continue about proving that the explanation of Piaget about child's development is a theory. Other constructs in Piaget's theory are the stages of intellectual development: sensorimotor, preoperational, concrete operations and formal operations.

What is the law that relates the stages of intellectual development? First, let's define a law. A law is a generalization about the causal, sequential, or other relationship between two or more constructs. "Piaget proposes the law that these constructs are related to each other as an invariant sequence: the sensorimotor stage is always followed by the preoperational stage; the preoperational stage is always followed by the concrete operations stage and the concrete operations stage is always followed by the formal operations stage".

Uses of theory

Theories serve several purposes. First they identify commonalities in isolated phenomena. For example Piaget identify the effects of sensorimotor intelligence on many infant behaviors. Theoretical constructs identify the universals of experience in order to make sense of that experience. Second the laws of a theory allow us to make prediction and to control phenomena. For example certain astronomic laws enable astronomers to make predictions about eclipses and other phenomena in the universe. The laws of a theory of learning allow special educators to make interventions that lead to positive changes in student behavior.

Approaches to theory development

The two main approaches of theory development are: grounded theory and scientific method. In grounded theory the constructs and laws are grounded in the set of data collected. In other words the constructs and laws dreved from the immediate set of data collected. The usefulness of constructs and laws are tested in a subsequent research.

The other approach is called scientific method. It consists by formulating a theory and then testing it by collecting data. The process unfolds in three steps.

1. One formulates a hypothesis
2. One makes deductions of observable consequences of the hypothesis.
3. One tests the hypothesis by collecting data.

Example of Theory testing

The three steps of theory testing are demonstrated in a study of self-attention theory led by Brian Mullen. This theory focuses on self-regulation processes that occur when an individual projects his attention on himself. Some manifestations of these processes are self-consciousness and embarrassment at work. One of the functions of self-theory is to explain the effects of groups on individuals. The theory states in part that when individuals are in groups they become more self-attentive as the size of the group decreases. This can be explained by the fact that when the size of the becomes smaller individuals can focus their attention more on themselves in relation to the group and therefore tend to follow the standards of the group.

The first step to test the validity of this theory is to formulate a hypothesis, which is a tentative proposition about the relationship between two or more theoretical constructs. "In Mullen's study, the hypothesis is that individuals would be more self-attentive in smaller groups than in larger groups" (Educational Research, an Introduction). The two theoretical constructs stated in the hypothesis are group size and self-attention. They are formulated in the hypothesis in inverse relation  to each other meaning that as group size decreases self-attention increases,

The second step in testing theory is to make deductions of observable consequences of the hypothesis. This process of deducting requires the existence of a real or simulated situation. In this perspective Mullen was able to obtain transcripts of 27 high school discussions of which size varies. Self-attention was operationally defined as the multiple uses of first person singular pronouns (I,me) by students when they talked in the discussion groups. The measure of group size was done by counting the number of students in each discussion group. Mullen was able to define and measure each construct stated in the hypothesis by using the available data.

The third step in testing a theory is to collect empirical data and determine whether they support or reject the hypothesis. Mullen counted the number of students and the number of first singular pronouns stated by the students in each discussion group using all the data available to him. Mullen uses the correlation method to show the relationship between the two sets of data obtained by counting the number of students and the number of pronouns. The result of the statistical analysis was: a higher frequency of first-person singular pronouns was encountered in the smaller groups than in the higher ones.

The hypothesis was supported by available data. Therefore that part of the theory corresponding to the is hypothesis is reinforced. This increases confidence that the theory provides a valid explanation on how people act in social situations.

Several weaknesses arise from the scientific method in spite of its power to test hypothesis.  "One of them is that the researcher may deduce inappropriate observable consequences from the hypothesis, and thus make an inappropriate testing of the hypothesis".

The other weakness is very difficult to overcome. "Any observable result potentially can support multiple, sometimes conflicting theories".  Therefore a researcher can never prove a theory but can only support it.