The number of cigarettes smoked, drugs taken or drinks imbibed would be the dependent variable in a study of stress and use of smoking, drinking or drug-taking. In fact there is a difference between personality factors, some being more resilient and less likely to turn to substances to relieve stress.
Although job stress has a positive and statistically significant impact on smoking intensity, this is truer for light smokers, while it has a positive and significant impact on alcohol consumption mainly for heavy drinkers according to a recent Canadian study. In short, it is not a simple case of finding that if you increase a person's stress, they will use substances more.
This is one of these cases where publication will hardly surprise the reader. Studies of the effect of drinking alcohol upon drivers are unlikely to pronounce that alcohol is safe to take before driving and has no effect on proficiency. The independent variable is alcohol use, the dependent variable is driving ability.
In a 2014 study, it was found, unsurprisingly, that under the influence of alcohol, most drivers tend to be more impulsive and adventurous and their abilities of judgment, vigilance, recognition, reaction, and controlling were impaired. The greater the amount of alcohol, the more reliable the findings. The correlation between drinking, driving and road accidents is nothing new.
A laboratory experiment need not take place in a formica sided area. It has to do with the control an experimenter is taking over the study. An experiment is an investigation in which a hypothesis is scientifically tested. In an experiment, an independent variable (the cause) is manipulated and the dependent variable (the effect) is measured; any extraneous variables are controlled.
This has to be so that experiments are objective. The views and opinions of the researcher should not affect the results of a study. The aim is to make the data more valid, and avoid contamination by interfering variables, such as distractions for the subjects, and from bias.
If a company wishes to know about the amount of something, whether this is a product, a type of employee behaviour or even a category of employee (call this E) it follows a procedure of counting. If the company is interested in the degree to which the product or employee represents E then individual products or persons must be ranked as first (having the most) second, and so on.
To discuss this means using the level of measurement termed 'ordinal'. There are four levels of measurement: nominal, ordinal, interval, and ratio.
An experimenter has an idea that a certain outcome will follow from a premise. It is usual that in setting up an experiment to prove this, the experimenter composes his hypothesis suggesting the opposite of what he believes, and then sets out to disprove that hypothesis.
This would be the hypothesis that there is no significant difference between specified populations, any observed difference being due to sampling or experimental error - yet with the hope, and probably expectation that the reverse will prove to be the case. The aim is to make a precise ultimate statement by reference to the statistical procedures and results obtained.