The firststepin the design of research is the selection of a question
that has its roots in an interesting theoretical argument. The range of
potential topics for social research is as broad as the range of social
behavior, and although this fact does not make it easier for the
researcher to make a choice, it does represent a very large universe of
ideas to explore.
It will help the decision makers evaluate your project. It’s critical that you have manageable objectives. (Two or three clear goals will help to keep your research project focused and relevant.)
In order to do this, you have to identify the environmental variables that will affect the research project.
If you understand the nature of the problem as a researcher, you will be able to better develop a solution for the problem.
To help you understand all dimensions, you might want to consider focus groups of consumers, sales people, managers, or professionals to provide what is sometimes much needed insight.
Such programs create a commitment to follow some behavioral pattern in the future.
Studying such a process involves:
Types of research designs
Historical Research Design - The purpose is to collect, verify, synthesize evidence to establish facts that defend or refute your hypothesis. It uses primary sources, secondary sources, and lots of qualitative data sources such as logs, diaries, official records, reports, etc. The limitation is that the sources must be both authentic and valid.
Case and Field Research Design - Also called ethnographic research, it uses direct observation to give a complete snapshot of a case that is being studied. It is useful when not much is known about a phenomenon. Uses few subjects.
Descriptive or Survey Research Design - It attempts to describe and explain conditions of the present by using many subjects and questionnaires to fully describe a phenomenon. Survey research design /survey methodology is one of the most popular for dissertation research. There are many advantages
.
Correlational or Prospective Research Design - It attempts to explore relationships to make predictions. It uses one set of subjects with two or more variables for each.
Causal Comparative or Ex Post Facto Research Design - This research design attempts to explore cause and affect relationships where causes already exist and cannot be manipulated. It uses what already exists and looks backward to explain why.
Developmental or Time Series Research Design - Data are collected at certain points in time going forward. There is an emphasis on time patterns and longitudinal growth or change.
Experimental Research Design - This design is most appropriate in controlled settings such as laboratories. The design assumes random assignment of subjects and random assignment to groups (E and C). It attempts to explore cause and affect relationships where causes can be manipulated to produce different kinds of effects. Because of the requirement of random assignment, this design can be difficult to execute in the real world (non laboratory) setting.
Quasi Experimental Research Design - This research design approximates the experimental design but does not have a control group. There is more error possible in the results.
5 Ways to Formulate the Research Problem
1. Specify the Research Objectives
A clear statement of objectives will help you develop effective research.It will help the decision makers evaluate your project. It’s critical that you have manageable objectives. (Two or three clear goals will help to keep your research project focused and relevant.)
2. Review the Environment or Context of the Research Problem
As a marketing researcher, you must work closely with your team. This will help you determine whether the findings of your project will produce enough information to be worth the cost.In order to do this, you have to identify the environmental variables that will affect the research project.
3. Explore the Nature of the Problem
Research problems range from simple to complex, depending on the number of variables and the nature of their relationship.If you understand the nature of the problem as a researcher, you will be able to better develop a solution for the problem.
To help you understand all dimensions, you might want to consider focus groups of consumers, sales people, managers, or professionals to provide what is sometimes much needed insight.
4. Define the Variable Relationships
Marketing plans often focus on creating a sequence of behaviors that occur over time, as in the adoption of a new package design, or the introduction of a new product.Such programs create a commitment to follow some behavioral pattern in the future.
Studying such a process involves:
- Determining which variables affect the solution to the problem.
- Determining the degree to which each variable can be controlled.
- Determining the functional relationships between the variables and which variables are critical to the solution of the problem.
5. The Consequences of Alternative Courses of Action
There are always consequences to any course of action. Anticipating and communicating the possible outcomes of various courses of action is a primary responsibility in the research process.Types of research designs
Historical Research Design - The purpose is to collect, verify, synthesize evidence to establish facts that defend or refute your hypothesis. It uses primary sources, secondary sources, and lots of qualitative data sources such as logs, diaries, official records, reports, etc. The limitation is that the sources must be both authentic and valid.
Case and Field Research Design - Also called ethnographic research, it uses direct observation to give a complete snapshot of a case that is being studied. It is useful when not much is known about a phenomenon. Uses few subjects.
Descriptive or Survey Research Design - It attempts to describe and explain conditions of the present by using many subjects and questionnaires to fully describe a phenomenon. Survey research design /survey methodology is one of the most popular for dissertation research. There are many advantages
.
Correlational or Prospective Research Design - It attempts to explore relationships to make predictions. It uses one set of subjects with two or more variables for each.
Causal Comparative or Ex Post Facto Research Design - This research design attempts to explore cause and affect relationships where causes already exist and cannot be manipulated. It uses what already exists and looks backward to explain why.
Developmental or Time Series Research Design - Data are collected at certain points in time going forward. There is an emphasis on time patterns and longitudinal growth or change.
Experimental Research Design - This design is most appropriate in controlled settings such as laboratories. The design assumes random assignment of subjects and random assignment to groups (E and C). It attempts to explore cause and affect relationships where causes can be manipulated to produce different kinds of effects. Because of the requirement of random assignment, this design can be difficult to execute in the real world (non laboratory) setting.
Quasi Experimental Research Design - This research design approximates the experimental design but does not have a control group. There is more error possible in the results.
EXPERIMENTAL ERROR *
Experimental error – a fact of scientific life.
Experimental error is always with us; it is in the nature of scientific measurement that uncertainty
is associated with every quantitative result. This may be due to
inherent limitations in the measuring equipment, or of the measuring
techniques, or perhaps the experience and skill of the experimenter.
However mistakes do not count as part of the analysis, though it has to
be said that some of the accounts given by students dwell too often
on mistakes – blunders, let's not be coy – and too seldom on the
quantitative assessment of error. Perhaps it's easier to do so, but
it is not quantitative and does not present much of a test of the
quality of the results.
The development of the skill of error assessment is the purpose of these pages. They are not intended as a course in statistics, so there is nothing concerning the analysis of large amounts of data. |
The Origin
|
Errors – or uncertainties in experimental data – can arise in
numerous ways. Their quantitative assessment is necessary since only
then can a hypothesis be tested properly. The modern theory of atomic
structure is believed because it quantitatively predicted
all sorts of atomic properties; yet the experiments used to determine
them were inevitably subject to uncertainty, so that there has to be
some set of criteria that can be used to decide whether two compared
quantities are the same or not, or whether a particular reading
truly belongs to a set of readings. Melting point results from a
given set of trials is an example of the latter.
Blunders (mistakes).
Mistakes (or the much stronger 'blunder') such as, dropping a small amount of solid on the balance pan, are not errors in the sense meant in these pages.
Unfortunately many critiques of investigations written by students are fond of quoting blunders as a source of error, probably because they're easy to think of. They are neither quantitative nor helpful; experimental error in the true sense of uncertainty cannot be assessed if the experimenter was simply unskilled.
Mistakes (or the much stronger 'blunder') such as, dropping a small amount of solid on the balance pan, are not errors in the sense meant in these pages.
Unfortunately many critiques of investigations written by students are fond of quoting blunders as a source of error, probably because they're easy to think of. They are neither quantitative nor helpful; experimental error in the true sense of uncertainty cannot be assessed if the experimenter was simply unskilled.
Human error.
This is often confused with blunders, but is rather different – though one person's human error is another's blunder, no doubt. Really it hinges on the experimenter doing the experiment truly to the best of his ability, but being let down by inexperience. Such errors lessen with practice. They also do not help in the quantitative assessment of error. An example of this would be transferring solids from the weighing boats to a test tube
Only if the human error has a significant impact on the experiment should the student mention it.
Instrumental limitations.
Uncertainties are inherent in any measuring instrument. A ruler, even if as well-made as is technologically possible, has calibrations of finite width; a 25.0 cm3 pipette of grade B accuracy delivers this volume to within 0.06 cm3 if used correctly. A digital balance showing three decimal places can only weigh to within 0.0005 g by its very nature and even then only if it rounds the figures to those three places.
Calibrations are made under certain conditions, which have to be reproduced if the calibrations are to be true within the specified limits. Volumetric apparatus is usually calibrated for 20oC, for example; the laboratory is usually at some other temperature.
Analogue devices such as thermometers or pipettes often require the observer to interpolate between graduations on the scale. Some people will be better at this than others.
These limitations exist and are unlikely significant errors in your experiment.
This is often confused with blunders, but is rather different – though one person's human error is another's blunder, no doubt. Really it hinges on the experimenter doing the experiment truly to the best of his ability, but being let down by inexperience. Such errors lessen with practice. They also do not help in the quantitative assessment of error. An example of this would be transferring solids from the weighing boats to a test tube
Only if the human error has a significant impact on the experiment should the student mention it.
Instrumental limitations.
Uncertainties are inherent in any measuring instrument. A ruler, even if as well-made as is technologically possible, has calibrations of finite width; a 25.0 cm3 pipette of grade B accuracy delivers this volume to within 0.06 cm3 if used correctly. A digital balance showing three decimal places can only weigh to within 0.0005 g by its very nature and even then only if it rounds the figures to those three places.
Calibrations are made under certain conditions, which have to be reproduced if the calibrations are to be true within the specified limits. Volumetric apparatus is usually calibrated for 20oC, for example; the laboratory is usually at some other temperature.
Analogue devices such as thermometers or pipettes often require the observer to interpolate between graduations on the scale. Some people will be better at this than others.
These limitations exist and are unlikely significant errors in your experiment.
Observing the system may cause errors.
If you have a hot liquid and you need to measure its temperature, you will dip a thermometer into it. This will inevitably cool the liquid slightly. The amount of cooling is unlikely to be a source of major error, but it is there nevertheless.
If you have a hot liquid and you need to measure its temperature, you will dip a thermometer into it. This will inevitably cool the liquid slightly. The amount of cooling is unlikely to be a source of major error, but it is there nevertheless.
Errors due to external influences.
Such errors may come from draughts on the balance pan, for example (though this seems pretty close to a blunder), or maybe from impurity in the chemicals used. Again such things are unlikely to be significant in a carefully-designed and executed experiment, but are often discussed by students, again because they are fairly obvious things.
Such errors may come from draughts on the balance pan, for example (though this seems pretty close to a blunder), or maybe from impurity in the chemicals used. Again such things are unlikely to be significant in a carefully-designed and executed experiment, but are often discussed by students, again because they are fairly obvious things.
Not all measurements have well-defined values.
The temperature of a system, or its mass, for example, has particular values which can be determined to acceptable degrees of uncertainty with suitable care. Other properties do not; the diameter of a planet, for example, although quoted in tables of data, is a mean value. The same is true for the thickness of a piece of paper or the diameter of a wire. These measurements will vary somewhat at different places. It is important to realize what sort of data you are dealing with.
The temperature of a system, or its mass, for example, has particular values which can be determined to acceptable degrees of uncertainty with suitable care. Other properties do not; the diameter of a planet, for example, although quoted in tables of data, is a mean value. The same is true for the thickness of a piece of paper or the diameter of a wire. These measurements will vary somewhat at different places. It is important to realize what sort of data you are dealing with.
Sampling.
Many scientific measurements are made on populations. Such as final value that you report for melting point is from a population, albeit rather a small one. It is intuitively understood that the more samples you have from a given population the less the error is likely to be. It is why students shouldn’t be satisfied with one melting point of a substance, but should obtain at least two melting points.
Related to this are errors arising from unrepresentative samples. Suppose that a chemist wishes to time a particular reaction in a certain hood that is situated near a drafty vent in lab. The rate of this reaction will depend on how drafty that area, if the heating or cooling is on, the ambient temperature of the lab during busy and slow periods etc. So a measurement made at 3 o'clock on a Friday afternoon may be utterly unrepresentative of the mean rate of the reaction at some other location in lab or time period. It doesn't matter how many samples one takes – if the sampling method is this biased, a true picture cannot be obtained. Therefore a large sampling does not of itself ensure greater accuracy.
The bias in this example is fairly obvious. This is not always so, even to experienced investigators. Sir Ronald Fisher's famous text 'The Design of Experiments' deals with the difficulties of removing bias in biological investigations, and is a work on statistical methods. Although this degree of analysis may seem outside of our realm of experimental work, it will not be so if you go on to do research in many fields of science.
Many scientific measurements are made on populations. Such as final value that you report for melting point is from a population, albeit rather a small one. It is intuitively understood that the more samples you have from a given population the less the error is likely to be. It is why students shouldn’t be satisfied with one melting point of a substance, but should obtain at least two melting points.
Related to this are errors arising from unrepresentative samples. Suppose that a chemist wishes to time a particular reaction in a certain hood that is situated near a drafty vent in lab. The rate of this reaction will depend on how drafty that area, if the heating or cooling is on, the ambient temperature of the lab during busy and slow periods etc. So a measurement made at 3 o'clock on a Friday afternoon may be utterly unrepresentative of the mean rate of the reaction at some other location in lab or time period. It doesn't matter how many samples one takes – if the sampling method is this biased, a true picture cannot be obtained. Therefore a large sampling does not of itself ensure greater accuracy.
The bias in this example is fairly obvious. This is not always so, even to experienced investigators. Sir Ronald Fisher's famous text 'The Design of Experiments' deals with the difficulties of removing bias in biological investigations, and is a work on statistical methods. Although this degree of analysis may seem outside of our realm of experimental work, it will not be so if you go on to do research in many fields of science.
