Many events happen without any scientific observer on the sidelines.
If each sea monster broke water before a panel of ichthyologists, and each revolution were staged before a team of visiting sociologists, our knowledge would be far more complete. But for many phenomena the only reports we have are the casual impressions of untrained observers who happened to be there; these reports may be interesting and possibly useful, but must be interpreted most cautiously by scientists. Scientific observations are made under controlled conditions. (Experimentation) Laboratories are popular with scientists because they are handy places to control variables such as heat, light, air pressure, time intervals, or whatever is important. A variable is anything which varies from case to case. For example, people vary in height, weight, age, sex, race, religion, education, occupation, income, health, behavior characteristics, and many other things. We have a scientific experiment when we control all important variables except one, then see what happens when that one is varied. Unless all variables except one have been controlled, we cannot be sure which variable has produced the results. If we wish to study, say, the effects of phosphates on plant growth, all other factors—seed, soil, water, sunlight, temperature, humidity—must be the same for all the sample plots; then the varying amounts of phosphates on different test plots can be held responsible for different growth rates. This is the basic technique in all scientific experimentation—allows one variable to vary while holding all other variables constant.

Complicated statistical procedures for multivariate analysis
These enable the researcher to work with two or more variables at a time. But this is only a refinement of the basic procedure of holding all other variables constant in order to measure the impact of the one (or more) under study. Failure to control all variables is a most common error in scientific method and ac-counts for most false conclusions. For example, a number of studies several decades ago concluded that the employment of mothers increased child delinquency. But these studies failed to control for the variables of social class and family composition. A sample of working mothers who were mostly poor, uneducated, often widowed or separated, and living in wretched neighborhoods, was compared with a sample of nonworking mothers who were more prosperous, better educated, and living with their husbands in good neighborhoods. No wonder they found strong association between maternal employment and child delinquency which more re-cent research does not fully confirm. Failure to control some variable—social class, education, age, and occupation are common ones—has invalidated many research studies. Since laboratories are such convenient places to control the conditions of observation, scientists use them whenever possible. But much that is important cannot be dragged into a laboratory. Volcanoes and earthquakes cannot be staged in a test tube, nor can we study the courtship process very realistically by herding couples into a laboratory. Both physical and social scientists frequently must observe phenomena in their natural setting. Techniques may range from lowering a bathysphere to the ocean floor to giving a questionnaire to a group of army recruits. If we remember that the basic scientific procedure is the conducting of accurate observations, while laboratories, instruments, arid computers are merely tools of observation, this difference in technique will not confuse us.
The scientific critics will trust a reported observation only insofar as the conditions of observation have been controlled. On this basis scientists are skeptical of the claims of spiritualism and mind reading. Spiritualists can conduct a very convincing scene in their own stage setting but are loath to attempt a séance where the room, furnishings, and lighting are controlled by the scientist. The professional mind reader is very convincing in a theater setting but is unwilling to attempt a reading under scientifically controlled conditions. Until spiritualists and mind readers make demonstrations under conditions which preclude the possibility of deception, scientists should dismiss them as either entertainers or frauds. Is it not strange that most of those who claim to foresee the future are performing in shabby carnivals and nightclubs instead of raking in millions in Wall Street? Why has no professed mind reader ever won a world chess or bridge championship? Although there are occasional newspaper reports of some psychic having "solved" a crime, is it not significant that police departments and intelligence agencies do not routinely employ psychic detectives? Although many exposes of the tricks of psychics, mentalists, fortune tellers, astrologers, and spiritualists have been published, their popular following today seems greater than at any time in recent history. In these several respects, then, scientific observation differs from looking at things. We spend our lives looking at things, and this activity brings us much information, many impressions, and numerous conclusions. But these conclusions are clouded by accident of coincidence, by selective memory, and by personal bias. Therefore, before accepting any generalization as true, the critical observer wants to know what it is based upon. A conclusion is based upon a systematically collected body of scientific evidence, or it is an offhand reaction to haphazard observation.

Do you like this article? Share It!