Facebook Pixel Code
Diagnostic imaging
Diagnostic Imaging

Conducting Research

Designing a Study

The scientific approach to research is a method of using unbiased observation to form and test beliefs.  Using a scientific approach involves:

  • Raising a fundamental question of interest that is addressable using scientific research
  • Researching what is already known about this fundamental question
  • Make a prediction or hypothesis in answer to the question of interest
  • Plan and implement an experiment to test the prediction
  • Reflect on the results of the experiment and how they affect what was known before.
  • Be alert for how the new data does or does not readily fit into the existing structure of scientific understanding
  • Communicate new knowledge by participating in oral presentations and submitting papers. 
  •  As the scientific community judges the validity and importance of the results, new questions will be raised.

Where do ideas (Hypotheses) come from?

  • They may expand on previous research
  • They could enhance current diagnostic processes/drugs
  • They may be from specific requests from experts in the field/community
  • They may arise with the emergence of new diseases or new diagnostic techniques
  • The may arise due to new discoveries or incidental findings

Hypotheses

Definition from the On-Line Merriam-Webster Dictionary - 1(a): an assumption or concession made for the sake of argument 1(b): an interpretation of a practical situation or condition taken as the ground for action 2: a tentative assumption made in order to draw out and test its logical or empirical consequences 3: the antecedent clause of a conditional statement

Types of Hypotheses:

Experimental Hypothesis: prediction that treatment (independant variable) will cause a particular effect in dependant variable.  An experiment attempts to support experimental hypothesis, so we can reject null hypothesis OR A prediction that a treatment will CAUSE an effect.

Null Hypothesis: a statistical hypothesis to be tested and accepted or rejected in favor of an alternative; specifically : the hypothesis that an observed difference (as between the means of two samples) is due to chance alone and not due to a systematic cause OR A prediction that the treatment will NOT have an affect (any noted differences could be due to chance).  If you are able to reject a null hypothesis then you can skeptically accept the hypothesis that the treatment had an effect.  If you fail to reject the null hypothesis you CANNOT draw any conclusions.

Definition of Experiment

Definition from the On-Line Merriam-Webster Dictionary: A trial or special observation, made to confirm or disprove something doubtful; esp., one under conditions determined by the experimenter; an act or operation undertaken in order to discover some unknown principle or effect, or to test, establish, or illustrate some suggest or known truth; practical test; poof..

Definition of Randomization 

Randomization is a core principle in the statistical theory of design of experiments. Randomization is not haphazard; it is understood that randomization reduces bias by equalizing other factors that have not been explicitly accounted for in the experimental design and because it produces ignorable designs. In design of experiments, most statisticians prefer Completely Randomized Designs. Other experimental designs are used when a full randomization is not possible. These cases include experiments that involve blocking and experiments that have hard-to-change factors.

Definitions of Variables

A variable is a characteristic of a person, object or phenomenon which can take on different values. These may be in the form of numbers (e.g., age) or non-numerical characteristics (e.g., sex). Variables aren't always 'quantitative' or numerical. 

Numerical variables can either be continuous or discrete.

  1. Continuous. With this type of data, one can develop more and more accurate measurements depending on the instrument used, e.g.:
    • height in centimetres (2.5 cm or 2.546 cm or 2.543216 cm)
    • temperature in degrees Celsius (37.20C or 37.199990C etc.)
  2. Discrete. These are variables in which numbers can only have full values, e.g.:
    • number of visits to a clinic (0, 1, 2, 3, 4, etc).
    • number of CT scans performed (0, 1, 2, 3, 4, 5, etc.)

Categorical variables, on the other hand, can either be ordinal or nominal.

  1. Ordinal variables. These are grouped variables that are ordered or ranked in increasing or decreasing order.
  2. Nominal variables. The groups in these variables do not have an order or ranking in them.

An important distinction having to do with the term 'variable' is the distinction between an independent and dependent variable.

Independent Variable: An independent variable can be manipulated by the researcher

Dependent Variable: Dependent variables are not manipulated by the researcher.  Changes in the dependent variables can be due to chance, outside factors, or by a treatment.

An attribute is a specific value on a variable. For instance, the variable sex or gender has two attributes: male and female. This distinction is particularly relevant when you are investigating cause-effect relationships. 

Research Design

Even the best statistics are useless if the research design has not accounted for possible BIAS.

A Randomized Controlled Trial (RCT) is the strongest research design.  The participants/subjects must be randomly placed in ¡Ý2 study groups.  Advantages to a RCT are that they are the strongest research design and that it prevents systematic differences between groups since they are randomly selected from the same population.  Disadvantages to RCT are that they are the most expensive types of trials and subjects who may volunteer may not be representative of patients in general.

Cohort Studies are studies where a group shares something similar.  Advantages to a Cohort study is that it is cheaper than RCTs, more feasible and usually have few ethical issues.  Disadvantages to Cohort studies are you cannot control other factors (variables) making it more difficult to prove cause and effect, or if a treatment becomes popular it becomes difficult to find a cohort that has not received the treatment, and it is nearly impossible to have a blinded study.

Cross Over Trial Design is where the subjects get both treatments