When assessing research results, high quality standards have to be applied. Methodological failings such as missing data on exposure or the state of the animal or subject, as well as missing control experiments complicate the analysis and interpretation of research results regarding potential health effects on human beings.

The Federal Office for Radiation Protection (BfS) continually tracks and evaluates relevant international and national publications in accredited scientific journals based on scientific criteria. As a general rule, research will be considered which has been "peer-reviewed“ prior to publication in scientific journals. This largely guarantees that research meets fundamental quality standards. Nonetheless, extreme quality differences can occur. Therefore, when evaluating the research design (design of experiment), the testing process and the analysis of the results, the Federal Office for Radiation Protection attaches great importance to comprehensive and comprehensible information, especially regarding exposure conditions, accompanying checks and statements of results analysis.

Exposure conditions

Indispensable prerequisites to assess the quality of the presented project are detailed information on parameters and conditions of exposure, such as frequency, intensity, orientation to the antenna, as well as the methods determining these values. Especially during animal testing the power actually absorbed (measured in Watt per kilogram) by the tissue can vary significantly, depending on test conditions. Several factors, such as technical prerequisites of the exposure facility as well as the mobility of the animals (free moving or fixed animals) play an important role. Every inaccuracy in this respect will inevitably lead to misinterpretation of research results.

Controls

Every scientific experiment is monitored with controls. Adequate controls exclude alternate causes for the observed effects as much as possible. For example, when tests with mice or rats are carried out in order to specifically expose the head to high frequency electromagnetic fields, the animals are often restrained. These restraints can be stressful for the animals. In order to be able to distinguish between effects caused by stress or by exposure free-moving animals (cage control) are compared with animals inside the exposure facility but not exposed to electromagnetic fields (sham-exposure) and to those who were actually exposed. Naturally, control animals and exposed animals should not vary much in other respects (age, size, sex, weight etc.). Additionally in many scientific experiments, so called positive controls are included. These animals are usually treated with substances which will guarantee to cause the investigated effects e.g. the development of DNA strand breaks. Therefore, positive controls serve to test the chosen research method.

„Blinded“ study design

A further important standard is the so-called “blinded“ study design. This means, for example in animal studies that the persons conducting, and in particular the persons analysing the experiments do not know whether they are dealing with a control or an exposed animal. This procedure aims at minimising the risk that the analysis is biased by personal expectations. These expectations can be negative or positive - both cases may lead to a subjective analysis, which may skew the real results.

This concept is particularly essential for human studies (volunteer studies). It has to be considered here, that not only the person conducting the experiment has expectations, but also the volunteer, might it be conscious or unconscious, positive or negative. If the goal is to robustly investigate real correlations between exposure and effects and to exclude placebo-effects (belief in a positive effect) as well as nocebo-effects (belief in a damaging impact), emphasis must be placed on a “double-blind” design, meaning that neither the volunteer nor the person conducting the experiment know whether an exposure has occurred or not. The study is “unblinded” only after the analysis of the investigated endpoints and then the findings are assigned to the actual exposure status.