Construct is a variable that is not directly observable or measurable. But once a construct has been operationally defined, variables are created. Examples of Construct: effort, itchiness, hunger, maturity, wisdom...
| Construct | Operational definition |
|---|---|
Stress |
Level of cortisol (stress hormone) |
Hunger |
Gramms of food consumed |
Effort |
Minutes spent studying for an exam |
Operational definition describes how researcher decide to measure the variables (in our case construct) in a study. It also helps you to measure constructs in the real world by turning them into measurable variables
Hypothesis is a statementabout the relationship between the variables.
All experiments/researches examine some kind of variable(s). A variable is not only something that we measure, but also something that we can manipulate and something we can control for.
- Dependent Variable or Outcome, or y-variable.
- Is a variable that is dependent on an independent variable(s).
- Independent Variable sometimes called Experimental Variable or Manipulated Variable, or Predicted, or x-variable.
- Is a variable that is being manipulated in an experiment in order to observe the effect on a Dependent Variable, sometimes called an Outcome Variable.
- Lurking Variables or Extraneous factors are variables/factors that can impact the Outcome/Dependent Variable.
Population (or mu) are values that describe the entire population.
A parameter is any numerical quantity that characterizes a given population or some aspect of it. This means the parameter tells us something about the whole population. Example of parameters: standard deviation, population mean (average)
N is a population size.
mu is an average (or a mean) of the entire population.
Sample (or X-bar) are portions of a population selected for the study. A measurable characteristic of a sample is called a statistic.
n is a number of a sample.
X-bar is an sample average (or a mean) of the population.
Random sample means that each element in the population has an equal chance of being included to the sample.
Random selection (or sampling) is a randomly choosing a sample from a population.
Convenience selection (or sampling) selections is based on easy availability/accessibility of elements; doesn't represent entire population
Samplig error the difference between a population parameter and a sample statistic used to estimate it. Sampling error occurs because a portion, and not the entire population, is surveyed.
Sampling error formula:
mu - X-barorX-bar - muwheremuis a population average andX-baris a sample average
Bias - any systematic failure of a sample to represent its population. The most common is called a simple random bias. The best way to avoid random bias is to select elements for the sample at random. Non-response bias occurs when individuals randomly sampled for a survey fail to respond, cannot respond or decline to participate.
Links: