Multicenter (pilot) study of the efficacy of gliatilin in acute ischemic stroke. The Importance of Client Confidentiality

Topic 4. Basic terms used in evidence-based medicine.

With the introduction of the principles of evidence-based medicine into healthcare practice, the role of clinical drug trials has increased, since effective clinical decisions can be made only on the basis of well-planned, controlled clinical trials.

Currently, preference is given to such a design of clinical drug trials, which provides the most reliable data. For example, conducting prospective controlled comparative randomized and, preferably, double-blind studies.

So what can clinical research look like?

A case control study is usually a retrospective study that compares people with a particular disease or outcome (“case”) with people in the same population who do not have that disease or who do not experience that outcome (“ control") to identify associations between outcome and prior exposure to certain risk factors.

A cohort study is a study in which a group of people (cohort) identified according to certain characteristics is observed for some time. Outcomes or outcomes are compared for subjects in different subgroups of a given cohort, those who were exposed or not exposed (or were exposed to varying degrees) treatment with study drug.

In a prospective cohort study, cohorts are formed in the present and observed in the future.

In a retrospective (or historical) cohort study (Retrospective cohort study), a cohort is selected from archival records and traced their outcomes from then to the present.

A controlled study is any study that controls (and, where possible, minimizes or eliminates) potential sources of bias.

Longitudinal study - a long-term clinical study in which long-term periodic observation of the same individuals is carried out.

An international study is a study that takes place in several countries.

Multicentral study - a study conducted in accordance with a single protocol in several research centers.

Non-comparative study - a study in which the comparator drug is not used (see comparative study).

Observational study - a study in which the researcher observes any indicator or the relationship of several indicators without interfering with the situation.

A one-stage study is a variant of a descriptive study that examines the impact of risk factors on a population and/or the prevalence of a disease (condition) in it at a particular point in time.

Pivot study - a study conducted according to the rules of the CCP and carefully monitored to ensure the validity of the data. During the pivotal study, basic information on efficacy and safety is obtained for submission to official authorities. Studies that are not subject to intensive monitoring are considered supportive.

An open study is a study in which the distribution of patients into treatment groups is known to all and where all participants know which drug the patient is receiving. This could be a study that is investigating one or more treatments.

Parallel study - a study in which subjects in different groups receive either only the study drug, or only the comparison drug or placebo. In a parallel study, two or more groups of subjects are compared, one or more of whom receive an investigational drug and one group is a control. Some parallel studies compare different treatments without including a control group. Such a scheme is called a scheme of independent groups.

A cross-over study is a study in which each patient receives both drugs being compared, usually in random order. This regimen is used in patients with stable disease, usually chronic. Currently, such a study scheme is not widely used due to the emerging carryover effects when changing drugs.

A pilot study is a study conducted on a small number of patients and is intended to provide preliminary data that are important for planning further stages of the study.

Another goal of conducting a pilot study is to get an idea of the difference between the two treatments, which will allow to establish the number of patients that should be included in the main study (sample size) and the required power for exact definition this difference.

A placebo-controlled trial is a comparative study in which the activity of a new drug is compared to that of a placebo. It can be conducted as a parallel group study or as a crossover study. In the United States, two placebo-controlled trials are required for a new drug application.

A prospective study is a study in which patients are recruited according to the criteria set out in the study protocol. Patients receive an investigational drug and are observed for some time. That is, the formation of groups receiving or not receiving the investigational medicinal product occurs before the results are recorded. Most clinical studies are prospective.

Randomized study - a study in which patients are randomly assigned to treatment groups (used special procedure randomization) and have the same chance of receiving study or control product (comparator or placebo).

Retrospective (historical) study (Retrospective study) - a study in which the outcomes of previously conducted clinical trials or studies are studied, that is, the outcomes have already occurred before the study was started. The researcher reviews medical records and selects patients according to certain criteria in order to study the results of treatment.

Blind or masked study (Blind study) - a study when there is no information about which group - experimental or control - each subject is assigned to. There are several options for such studies:

simple blind study (blind study) - information is missing only for the patient;

double-blind study (double-blind study) - information is missing from both the patient and the researcher;

triple or absolute blind study (total-blind study) - information is missing from the patient, the researcher and the persons conducting statistical processing of the study results.

Masked studies are used to eliminate bias in clinical trials.

A comparative study is a study in which an investigational drug, the efficacy and safety of which is not yet fully understood, is compared with a drug whose efficacy and safety are well known. This may be placebo, standard therapy, or no treatment at all.

Stratified study - a study in which patients are divided into different groups (strata). Patients are assigned to strata according to certain parameters that are different from the parameters of patients in another stratum. The stratification can be prospective (control of enrollment is necessary to achieve a certain balance) or retrospective (eg, in the analysis).

An experimental study is a study whose conditions are under the direct control of the researcher. Most often, this is a comparative study, pre-planned and devoted to the study of the influence of at least one factor. Comparisons can be made between two or more groups or within the same group before and after the intervention.

- Basic terms (in alphabetical order)

Total number of observations

Numerical value N (total number of observations) = A+B+C+D

Absolute Benefit Increase (ABI): frequency difference favorable outcomes between experimental and control groups:

Absolute Risk Increase (ARI): the absolute difference in the frequencies of adverse outcomes, which are higher in the experimental group than in the control group; ARI is also used in assessing the impact of risk factors on disease.

Absolute Risk Reduction (ARR). Absolute risk reduction. The difference in the frequencies of adverse outcomes between the experimental and control groups:

best evidence. Reasonable proof. Basis for decision-making in health care. A well-designed and conducted randomized controlled trial is considered the best for evaluating the effectiveness of an intervention, but other types of trials may be more appropriate in some cases.

blind study. Blind research. One or more parties involved do not know how patients are assigned to treatment and control groups. With single blind (Single Blind) - the distribution is unknown to patients, in double blind (Double Blind) - to researchers and patients, in triple (Triple Blind) - also to the persons organizing the study and analyzing its results, and in full blind - everyone is in the dark who interacts with patients, investigator and data; and patient codes are prepared by people not associated with the study. Used to eliminate bias in clinical trials.

clinical trials. Clinical trials are a special type of research for the evaluation of various interventions, the conditions for which (selection of intervention groups, nature of the intervention, organization of observation and evaluation of outcomes) are aimed at eliminating the influence of systematic errors on the results obtained.

clinical epidemiology. Clinical epidemiology is a science that develops methods of clinical research that make it possible to draw fair conclusions by controlling the influence of systematic and random errors.

Comparator (Drug). Comparator drug. Placebo or active substance, the effectiveness and safety of which is known in advance.

Comparative Study. Comparative study. A study comparing a new drug to a known drug, placebo, or no treatment at all.

Compliance. Patient compliance with instructions for taking the study drug.

Confidence Interval (CI). Russian abbreviation - DI. Confidence interval. The 95% CI gives the limits within which, with 95% confidence, the values of the value under study are found for the patient population from which the study group is selected. With an increase in groups, the CI values narrow and the significance of differences increases. It is preferable to use CI instead of p, since CI characterizes the degree of confidence, and p only estimates the probability of no difference between groups (null hypothesis). This is a statistical indicator that allows you to assess the range within which the true value of a parameter in a population can lie; fluctuation range of true values. The values obtained in studies on a sample of patients differ from the true values in the population due to the influence of chance. So, in 95% of observations, the confidence interval means that the true value of the quantity with a probability of 95% lies within its limits. Confidence intervals help determine whether the given range values to the reader's ideas about the clinical significance of the effect and what results can be expected by applying the described technique on a similar group of patients.

continuous data. Continuous data is quantitative data that can take on any value on a continuous scale. Examples: body weight, arterial pressure, the ratio of partial pressure of oxygen to arterial blood).

control group. Control group (comparison group). The group of patients receiving during comparative study a drug whose efficacy and safety are well known (comparator drug). This may be placebo, standard therapy, or no treatment at all. The results of measurements in the control group are compared with the results of measurements in the experimental group to evaluate the effect of the study method of treatment.

Control Event Rate (CER). Russian abbreviation CHNBI. The frequency of adverse outcomes in the control group.

Controlled Study. Russian abbreviation - KI. controlled study. The term is used in two meanings:

1) as a synonym for a comparative (having a control group) study;

2) any study conducted in accordance with the protocol. cost-benefit. "Cost-benefit" - both the cost and the results of treatment are evaluated in monetary terms.

Cost Effectiveness. "Cost-effectiveness" - in monetary terms, the cost of treatment is estimated, and the results are considered as the achievement of certain tasks. In domestic literature, the term “price-efficiency” is often used, which is similar in meaning.

Cost Utility. Cost-benefit is the assessment of outcomes in terms of quality of life as the benefits of one treatment over another.

Cost Minimization. "Cost minimization" - compares the cost of equally effective types of treatment.

Cost-of-Iness. "Cost of the disease" - an estimate of the direct and indirect costs of treating a particular disease.

Crossover study. Cross study. A study in which each patient receives the compared drugs in turn in a random sequence. It is used in patients with a stable course of the disease, usually chronic. Currently, this design is not common due to carryover effects when switching drugs.

Crude Agreement: Prior consent. The proportion of hits between all raters. If there are only two of them, then the score is (A+D)/N.

decision sciences. The science of decision making. An area of methodological research related to understanding and improving decision making.

Design (of a Study). Study design. Method, scheme of the study.

discrete data. Discrete data is quantitative data that is expressed as whole numbers. For example, the number of migraine attacks a patient has during a month.

Dichotomous data. Dichotomous data - signs that can have only two opposite meanings (present - absent, yes - no, alive - dead).

ecological fallacy. Ecological error - a systematic error associated with the fact that the diseased individuals in the survey group may not actually experience the influence of risk factors. Occurs when a group is distinguished by common indirect evidence(living in the same territory, belonging to the same profession), and not by the presence of a risk factor in individual subjects (for example, by alcohol consumption).

efficiency. Efficiency. The ability of a drug to produce a desired therapeutic effect on the course of a disease.

Experimental Event Rate (EER). The frequency of adverse outcomes in the experimental group:

experimental group. An experimental group is a group that receives an intervention (treatment) during a study. Otherwise known as a treatment group or an intervention group (treatment group or Intervention group).

historical control. historical control. Comparison of those receiving investigational therapy with those previously treated with other therapy. The method is used when it is impossible to randomize patients in different groups. For example, in a study of long-term survival after kidney transplantation, it is impossible to randomize patients into "operating" and "non-surgical" groups, since it is unethical to deprive a person of the possibility of transplantation. The best solution there will be a comparison with those previously treated when there was no possibility of transplantation.

informed consent. Informed Consent. Signed by the study participants prior to the start of the RCT.

Kappa-criterion: (Greek κ - kappa). A goodness-of-fit test adjusted for a situation in which raters agree among themselves on a certain proportion of random events. If p is the observed proportion of cases of general agreement, and p is the proportion of supposed random events, then:

Likelihood Ratio for a Positive Test Result. Likelihood ratio for positive test results; The likelihood of a positive test result in a person with a disorder:

Likelihood Ratio for a Negative Test Result. Likelihood ratio for negative test results. The likelihood of a negative test result in a person with a disorder:

Long term trial. Long term study. Study of the efficacy and safety of the drug in long-term treatment diseases. The usual duration is more than 6 months. Often used when evaluating the impact of an intervention on hard endpoints.

Matched pair design. Scheme of the corresponding pairs (similar in meaning to "research in parallel groups"). A study plan in which the subjects of study are combined in pairs in accordance with certain factors (gender, age, etc.). One of them is receiving the investigational treatment, the other is receiving the comparator drug.

meta-analysis. Meta-Analysis. Methodology for combining heterogeneous studies related to the same topic by different authors in order to increase the reliability of estimates of similar results. Quantitative Analysis pooled results from multiple clinical trials of the same intervention. This approach provides more statistical power than in any single trial by increasing the sample size. Used to summarize the results of many tests. Cumulative meta-analysis is a method of calculating a generalized point estimate of the effect under study, in which studies are added one at a time in a certain order. Meta-regression analysis - regression analysis, where individual studies are considered as the unit of observation.

Negative Predictive Value. Negative predictive value. The ratio of the number of people with actual versus tested absence of the disorder.

Noncomparative Study. An incomparable study. A study in which the comparator drug is not used.

Numerical, or interval data. Quantitative data - quantities that are inherent in the natural order of arrangement with equal intervals between consecutive values, regardless of their place on the scale (body weight, glucose, cholesterol or creatinine in the blood).

Number Needed to Harm (NNH). Russian abbreviation - CHBNLNI. The number of patients in the experimental group compared with the control group required to identify an additional adverse outcome:

That is, to prevent one additional adverse outcome in the experimental group, there should be 30 patients.

Number Needed to Treat (NNT). The Russian abbreviation is CHBNLBI. The number of patients in the experimental group required to obtain an additional favorable outcome:

This indicator is the reciprocal of the absolute risk reduction.

This is a way of assessing the relative effectiveness of two treatments and shows how many patients need to be treated with the study method to prevent one occurrence of the outcome under study. For example, if 100 patients with arterial hypertension within 5 years, then NNT is 100 in 5 years.

Odds Ratio (OR). Russian abbreviation - OR. Odds ratio of outcome between groups of patients in a trial. The ratio of the probability that an event will occur to the probability that the event will not occur. Chances and probabilities contain the same information but express it differently. If the probability that an event will occur is denoted by P, then the chances of this event will be equal to P/(1-P). For example, if the probability of recovery is 0.3, then the chances of recovery are 0.3/(1-0.3) = 0.43. Chances are more convenient to use for some calculations than probabilities.

Used in systematic reviews or specific clinical trials. If, for example:

if a = 20 out of (a + b) = 100 experimental patients observed a certain outcome, and b = 80 did not, then the chances of this outcome in this group are: a/b = 20/80 = 0.25. The same for the control group at (c + d) = 100, c = 40 and d = 60 is c/d = 40/60 = 0.67 and the odds ratio is: OR = (a/b)/(c/d) = 0.25/0.67 = 0.37.

The odds ratio is similar to the relative risk, but numerically they are similar only for very low frequencies outcome. So, for this example, RR = (a/(a+b) : c/(c+d)) = 20/100: 40/100 = 0.50, and if the outcome rates were 1% and 2%, then RR value = 0.5 and OR = 1/99:2/98 = 0.49.

In case-control studies, the odds ratio is used to assess self-risk.

open study. Open study. A study where the distribution of patients into treatment groups is known to all. This could be a study that is investigating one or more treatments.

Ordinal data. Ordinal data - quantities that can be arranged in a natural order (ranked), for example, from small to large or from good to bad, but the size of the interval between such categories cannot be expressed quantitatively (for example: stages of the disease; ratings "high, medium , low" or "absent, weak, moderate, severe").

P value. The value of p is the probability that the result obtained is absolutely random. The value of p can vary from 1 (the result is definitely random) to 0 (the result is definitely not random). A p value less than or equal to a predetermined alpha error level (eg, 0.05) indicates that the difference obtained is statistically significant. At the same time, it does not at all reflect the clinical significance of the result.

Patient Expected Event Rate (PEER) is the patient-specific baseline risk relative to the average patient, expressed as decimal F, by which the value of NNT is divided. If exposure to adverse outcome F = 0.5 and NNT = 15, then:

pharmaceutical epidemiology. Pharmacoepidemiology. The science of applying epidemiological knowledge, methods and reasoning to the study of beneficial and adverse effects for the rational use of drugs in various human populations. The subject of the study are the results of observations collected in the daily clinical care of patients, in contrast to the refined data of randomized, double-blind clinical trials. The purpose of research is to determine the real effectiveness of drugs in clinical practice.

Pharmacoeconomics. Pharmacoeconomics. Cost analysis drug therapy for the healthcare system and society in order to identify the most beneficial drugs (treatments) for funders, patients and manufacturers. Decisions about financing plans for various health care programs, priorities among possible alternatives, and the patient's ability to use the best treatment option are based on the results of special (see below) types of pharmacoeconomic analysis.

placebo. placebo. A preparation (tablets, capsules, solutions) containing pharmacologically inactive components. Its effectiveness can only be explained psychological factor rather than an objective effect. If there is doubt about the safety of the components, they should not be included in the placebo. Used as a placebo dosage form, which is indistinguishable from the study drug in appearance, color, taste and smell, but has no specific effect (for example, glucose tablets or injections of isotonic sodium chloride solution), or other indifferent intervention used in medical research to simulate treatment in order to eliminate bias associated with the placebo effect.

placebo effect. Placebo effect - a change in the patient's condition (noted by the patient himself or by the attending physician), associated with the fact of treatment, and not biological action drug.

Placebo Controlled Trial. Placebo controlled study. The activity of the new drug is compared with placebo in a parallel group or crossover study.

protocol. Protocol. A document regulating the trial, including rationale, objectives, statistical analysis, methodology, conditions, inclusion and exclusion criteria for patients, etc.

positive predictive value. Positive predictive value. The ratio of the number of people who actually have a disorder to the number of people who test positive for the disorder. This value depends on the prevalence of the disorder:

Prevalence. Prevalence. The proportion of those suffering from this disorder in the study sample:

Prospective Study. prospective study. Patients are recruited according to protocol criteria, receive study drug, and are followed up to evaluate treatment outcomes. Most clinical studies are prospective.

Randomization. Randomization is a procedure that ensures the random distribution of patients into experimental and control groups. Random distribution achieves the absence of differences between the two groups and thus reduces the likelihood of bias in clinical trials due to group differences in any way.

relative risk. Relative risk or risk ratio (risk ratio) the ratio of incidence among individuals exposed and not exposed to risk factors. Relative risk does not carry information about the magnitude of the absolute risk (morbidity). Even with high relative risk values, the absolute risk may be very small if the disease is rare. Relative risk indicates the strength of the association between exposure and disease.

Relative Risk Reduction (RRR). Russian abbreviation - COR. Reduction of risk relative to the control group with indication of the confidence interval.

Relative Benefit Increase (RBI). Relative increase in the frequency of favorable outcomes in the experimental group compared to the control group:

Relative Risk Increase (RRI). Russian abbreviation - POR. The relative increase in the frequency of adverse outcomes in the experimental group compared to the control group is estimated in the same way as the RBI and is also used in assessing the impact of risk factors on the disease:

Randomized Clinical Study. Russian abbreviation RKI. Randomized clinical trial. Patients are assigned to treatment groups based on randomization and all have the same probability of receiving each of the drugs.

retrospective study. Retrospective study. A study in which a researcher reviews medical records and selects clinical records or published reports according to certain criteria in order to study treatment outcomes.

risk. Risk. The likelihood of harm to the health of the patient or causing him discomfort.

safety. Safety. One of key features drug (along with effectiveness), inversely proportional to the frequency of adverse drug reactions.

sensitivity. Test sensitivity. Proportion of those actually suffering from the disorder among those identified using the new test:

Sensitivity == 32/33 = 97.0%.

Short term Trial. Short term study. Study of the effectiveness of the drug in the short-term treatment of the disease. The usual duration is from 1 to 3 months. Often used in evaluating the pharmacodynamic effects of drugs.

specificity. Specificity. The ratio of the number of people who do not have the disorder to those who are found not to have the disorder using the new test:

Study/Trial. Clinical study/test. Grading system medical preparations devices or other type of intervention in a person (patient or volunteer) in order to identify any positive and negative effects that arose as a result of taking investigational drugs and / or studying their absorption, distribution, metabolism and excretion in order to evaluate the effectiveness and safety of drugs.

Surveillance. Drug surveillance. Collection of data on all adverse events associated with the use of drugs. As a rule, it is carried out for a long time.

survival analysis. Survival analysis. It is carried out at the end of the study in patients with a high risk of mortality. The need may arise in cardiac studies to assess survival after a heart attack.

Systematic error, bias. A systematic error is a non-random, unidirectional deviation of results from the true values. A systematic error can occur as a result of selection (when creating a sample) - sampling, or assembling bias, due to measurements - measurement bias, when exposed to unaccounted factors - confounding bias, and in many other cases. They also talk about a systematic error, meaning the bias in the publication of positive research results and the rejection of negative ones - publication bias. To combat systematic errors and obtain reliable data, organizational methods are used (for example, randomization, "blinding", etc.), as well as the introduction of corrections that take into account the magnitude of the bias.

Systematic Review. Systematized review. Summarization of primary data on one particular well-defined problem by standard reproducible methods using meta-analysis. In medicine, it is used to evaluate and compare therapeutic and side effects of treatment and make decisions about individual, subpopulation and population choice of treatment.

Three E "s Concept: Efficacy / Effectiveness / Efficiency. The concept of three E: efficiency / effectiveness / benefit. Benefit for the individual patient, patient populations and society as a whole.

validity. Suitability, suitability of the test for its intended purpose. There is no unambiguous criterion for suitability, since it can be defined in different ways. Reliability - a characteristic showing the extent to which the measurement result corresponds to the true value. The reliability of a study is determined by the extent to which the results obtained are valid for a given sample (internal validity). This is an internal characteristic and applies specifically to this group of patients and does not necessarily apply to other groups.

There are a lot of materials on the Internet on the topic of benefits and harms. palm oil For human body, but for some reason the majority sin with some one-sidedness, they don’t see complex problem, moreover, often contain common philistine myths. Let's figure it out.

subscribe to news

Pilot Study is intended to obtain preliminary data that is important for planning further stages of the study (determining the possibility of conducting a study in a larger number of subjects, the sample size in a future study, the required research power, etc.).

randomized clinical trial in which patients are assigned to treatment groups at random (randomization procedure) and have the same chance of receiving the study or control drug (comparator or placebo). In a non-randomized study, there is no randomization procedure.

Controlled (sometimes synonymous with "comparative") clinical trial in which an investigational drug whose efficacy and safety has not yet been fully established is compared with a drug whose efficacy and safety is well known (comparator drug). This may be placebo, standard therapy, or no treatment at all. IN uncontrolled (non-comparative) study the control/comparison group (the group of subjects taking the comparator drug) is not used. In a broader sense, controlled research refers to any research in which potential sources of bias are controlled (if possible, minimized or eliminated) (i.e., it is carried out in strict accordance with the protocol, monitored, etc.).

When conducting parallel studies subjects in different groups receive either the study drug alone or the comparator/placebo alone. IN cross studies each patient receives both compared drugs, usually in random order.

Research may be open when all participants in the study know which drug the patient is receiving, and blind (disguised) when one (single-blind study) or several parties participating in the study (double-blind, triple-blind, or full-blind study) are kept in the dark about the allocation of patients to treatment groups.

prospective study conducted by dividing the participants into groups who would or would not receive the study drug before the outcomes occurred. In contrast to him, in retrospective (historical) study the outcomes of previous clinical trials are being studied, i.e. outcomes occur before the study is started.

Depending on the number of research centers in which the study is conducted in accordance with a single protocol, studies are single-center and multicenter . If the study is conducted in several countries, it is called international.

In a parallel study, two or more groups of subjects are compared, one or more of whom receive an investigational drug and one group is a control. Some parallel studies compare different treatments without including a control group. (This design is called independent group design.)

cohort study is an observational study in which a selected group of people (cohort) is observed for some time. The outcomes of subjects in different subgroups of this cohort, those who were or were not treated (or were treated to varying degrees) with the study drug are compared. In a prospective cohort study, cohorts are formed in the present and observed in the future. In a retrospective (or historical) cohort study, a cohort is selected from archival records and traced through their outcomes from then to the present.

IN case-control study (synonym: similarity study) compare people with a particular disease or outcome (“case”) with people in the same population who do not have that disease or who do not experience that outcome (“control”) in order to identify an association between the outcome and prior exposure to certain risk factors. In a case series study, several individuals are observed, usually receiving the same treatment, without the use of a control group. The case report (synonyms: case report, medical history, single case description) is a study of treatment and outcome in one person.

Currently, preference is given to designing a clinical trial of drugs that provides the most reliable data, for example, when conducting prospective controlled comparative randomized and, preferably, double-blind studies.

IN Lately The role of clinical drug trials has increased due to the introduction of the principles of evidence-based medicine into practical healthcare. Chief among these is making specific clinical decisions for patient care based on the rigorously proven scientific evidence that can be obtained from well-designed, controlled clinical trials.

This chapter discusses a variety of field work that is carried out in customer satisfaction surveys. It touches upon such issues as “recruitment” and selection of interviewers, briefing, training, advanced training of interviewers and organization of their work. Some considerations are made about the degree of interest and involvement of the client company at this stage of the survey, the value of "listening sessions" during a telephone interview, and the problems that such "listening" and other forms of "accompaniment" during the interview can cause. . Issues of administrative management of field work are considered, including the need to control the correct use of the sample, as well as proposals for quality control. In addition to these issues, this chapter covers three additional topics. First, what can and should be done about consumers who raise serious issues during interviews that require immediate attention from the client company. Secondly, the issue of maintaining the confidentiality of the data obtained in the course of field work by the client. Finally, this chapter addresses the question of how to gain the trust and commitment of respondents.

Thank God, now, in the age of enlightenment and the establishment, everyone is entitled to their own opinion, and most often to the opinion that no one else has a right to it.

Ronald Knox

The author will not be surprised if many researchers, whether employed by companies or specialized agencies, choose to skip this chapter. After all, at this point, they believe their work is done. The requirements for the study were agreed, the questionnaire was approved and signed. All they have to do now is sit and wait for the interviewers to complete their part of the job. Such an approach is not only thoughtlessness, it shows almost complete absence respect and understanding in relation to the tasks facing both the respondent and the interviewer. In fact, the author believes that every researcher should work out some part of his career in the field by conducting interviews with respondents. If researchers did this, they would have a clearer idea not only of what the interviewer should do, but also of what to expect from the respondents.

Survey from the point of view of a potential respondent

Or another situation - a person has just got home after a busy day at work and a trip in a crowded train (or a whole hour of waiting in a traffic jam). He sat down for a few minutes to rest before taking the dog out for a walk or putting the kids to bed. But the phone rings and someone whose name he didn't quite catch starts asking him questions about whether he went to their bank the other day about an overdraft and what he thinks about the level of service he got there.

Or here: the person went on a business trip. As he settles into his room, he sees a small brochure on the table, which says that the hotel staff take pride in looking after the needs of those who come on a business trip, and would be very happy to know their opinion on the service. True, the review card looks like it has been sitting here for ten years and was used as a coffee tray by a previous guest.

What does it mean to be on the front line

According to the author, interviewers are real heroes who go out into the street in any weather and knock on the doors of houses, begging people to give an interview. How does it feel to be outside in January, on a cold, damp, and windy Saturday night, trying to complete a scheduled three interviews an hour? Why three? Because the administrator at the agency's headquarters calculated that it takes eight minutes to complete one interview, so the interviewer will spend less than thirty minutes of the allotted hour on three interviews, and the rest of the time will go from one door to another and find suitable respondents. Surely this administrator never went from house to house as an interviewer, so he did not calculate the time that goes into the usual greetings and exchanges of pleasantries, the need to remove outerwear and hang a coat somewhere, to unpack and plug in a laptop, to wait while the respondent makes you a cup of tea or coffee and offers you cookies. The administrator also did not take into account the mood of the interviewer, after the potential respondent angrily slammed the door in his face, because he was disturbed while watching an important football match on TV, or they were afraid to let him into the house just because it was raining, and from his coat water flows.

Interviewing can be a critical stage in a project and the impact the interviewer can have on the entire project cannot be underestimated. This is an area that both the client company and the agency researcher should pay close attention to.

Recruitment of interviewers

"Recruitment" and selection of interviewers is a very important part of the agency's work. What is he, an interviewer involved in market research? Many of them are employed part-time because they have other jobs or other responsibilities. Among the interviewers there are students who want to earn money for their studies, or actors who work part-time in the theatrical off-season, or able-bodied pensioners who are attracted not only by additional income, but also by the opportunity to fill their free time. But there are interviewers for whom the agency is a permanent place of work. Perhaps they like to work with people, but they do not sell anything at the same time. Some interviewers work for only one agency, but it is possible to work for several agencies at the same time.

A good interviewer should be friendly and well-mannered, organized, observant, have good communication skills and active listening techniques; among other things, he must love working with people. Those who conduct face-to-face interviews also need good health and a large supply of vitality, since they spend most of their time on their feet in the open. Interviewers should be able to count and have developed speaking skills. For international projects they need fluency foreign languages. Nowadays, more and more projects require the use of computers, because computerized personal and telephone surveys have gained great popularity. Therefore, many agencies strive to find interviewers who are not afraid to use new technologies.

case from practice

network research, a UK research agency specializing in telephone research on new product development and customer satisfaction, always conducts a preliminary telephone survey with potential interviewers as a starting point. The scope of the agency is telephone surveys, so if a candidate cannot make a favorable impression on the phone, he is unlikely to succeed in interviewing. Therefore, the intonation and timbre of the voice, as well as how he answers some simple questions, are of great importance. At the first meeting with a future interviewer network research also informs about the work schedule and payment terms. If the candidate is satisfied with this, he is asked about his experience in the field of market research and his ideas about the upcoming research. Experience is not prerequisite, because network research, like most agencies, provides intensive training programs to all interviewers, even if they have previously worked in this field. The reason for this is the reputation of the company (and the reputation of its customers). It is extremely important for a company that interviewers understand its culture and values and approach each survey in a consistent and professional manner.

Briefing and training of interviewers

In the UK, there are certain requirements for this category of workers, in particular the requirement that all interviewers must, in without fail receive appropriate training, and the sessions must be fully documented. For example, those interviewers who have already completed training must receive at least three hours of training when moving to a new agency, while newcomers attend a course of at least six hours. In addition, they perform their first task under the control and supervision of a mentor who checks their level of preparation. Documentary records are kept to confirm the date(s) and scope of the sessions conducted with the interviewers. In addition, all interviewers must be assessed at least once every six months.

The purpose of interviewing sessions is not only to mandatory requirement and inform them about safety at work and other administrative issues - payment procedures, lunch breaks, etc. All interviewer training should ensure a high standard of interviewing in any research project. During the sessions, interviewers should learn to introduce themselves and the project to the respondent, set the tone and pace, control the flow of the interview, ask leading questions in an open mind to clarify the respondent's answer, and work with and record respondents' free comments. Random comments are respondents' answers to open-ended questions, for example: why do you think so? It is extremely important to record such responses word for word or as closely as possible to the respondent's statement.

The ability to "manage" the respondent is a very important skill, especially when researching the level of customer satisfaction. When conducting a consumer survey on a subject such as the use of a product, the interviewer often encounters the fact that the respondent is not sufficiently knowledgeable about the research topic. But when it comes to the level of service or subjects that involve an emotional factor (for example, talking to men about cars), respondents can talk for a very long time. As a consequence, they often try to tell the interviewer much more about their experiences and opinions than the question on the questionnaire requires. The interviewer must very skillfully return the conversation to the subject of the survey and not alienate the respondent at the same time - as he may think that the interviewer is not really interested in what he is talking about, and he will simply lose the desire to answer subsequent questions.

Briefing Interviewers for a Customer Satisfaction Survey Project

In addition to general preparation, the interviewer has the right to expect a full briefing on each project on which he will work. Such coaching can take many forms. Personal coaching is the most expensive approach, in part because the interviewer will have to pay for travel to and from the class. But this is the most productive method of briefing, because the researcher can personally verify that all interviewers understand the instructions and know what is expected of them. In addition, the researcher can review all of the questions in the questionnaire and make sure that each of the interviewers understands the questions, the direction, and the quota requirements. In addition, in the classroom, an employee of the agency can answer any question of the interviewer.

If there is no opportunity to conduct personal classes - the project does not provide for expenses or the client is limited in funds - you can conduct a briefing by phone. But this method has its drawbacks. If you have to recruit big number respondents, the agency most often appoints group leaders who conduct telephone briefings with interviewers. To some extent, this increases the risk of misunderstanding or misrepresentation of information.

The least effective way of coaching is to send guidance materials by mail. At the same time, the agency naturally expects the interviewer to carefully read and understand the instructions, but there is no complete certainty about this, as well as whether the interviewer will remember everything he read when it is time to go out into the field.

Training and supervision during project implementation

How to keep the interest and high standard of the interviewer's work? This is not an easy task, especially on large-scale tracking projects, and especially when researching customer satisfaction. Questionnaires for projects of this type can be very monotonous. They often use the same response scale for all phases of a project (for good reason). Be that as it may, for the interviewer, the process can become very tedious. In addition, after a series of interviews of the same type, he gets the feeling that he has already "heard all this before." Attention begins to dissipate, the quality of work falls. In such a situation, control over the work of the interviewer becomes imperative. In the UK there is a data validation requirement for each survey. Real-time monitoring (via a remote listening device), listening to a tape-recorded interview, or making a phone call to the respondent after the interview can be used to check validity. In the UK, a minimum of 10% of all interviews in a survey must be supervised (unless the survey is conducted from a call centre, where 5% of the interviews are supervised).

Personality of the interviewer

It was previously mentioned that the personality of the interviewer is one of the factors that can affect the person's suitability for this role. People with a certain personality type may have difficulty meeting the necessary requirements.

A good interview is akin to an ordinary conversation between two people, but one of the interlocutors must strictly adhere to the “script” - it is very important to ask the respondents the same questions in the same manner. This is where difficulties arise, especially when it is necessary to record, word for word, arbitrary answers of respondents to open-ended questions. During briefing sessions, as well as during subsequent training, the agency trains the interviewer to ask leading questions if there is a need to clarify the respondent's answer or obtain additional information. But interviewers are, first of all, people, and everyone has their own characteristics. A person of choleric temperament may find it difficult to comply with the strict requirements of the agency. Can such interviewers be allowed to work on the project?

Take this situation as an example: many agencies specify specific questions that the interviewer should ask as a guide. Some of them might be, for example: “How interesting, could you tell us more about this?” or “But why do you think so?” However, in the mouths of some interviewers, such questions sound unnatural. More natural would be questions like “What, seriously?” or "Oh my God, really?" In most cases, the respondents react vividly to such comments, give more detailed and exhaustive answers. Can the interviewer be allowed to continue like this or be required to strictly follow the instructions?

According to the author, this is not the case when a bias is introduced into the survey, because the interviewer in this case does not put any ideas into the mind of the respondent. Perhaps, on the contrary, he receives from the respondent a response of higher quality (and, therefore, more valuable to the client) than when using the standard methodology. This does not mean that one can welcome anarchy. Each situation needs to be considered specifically. The author recalls how one British banker was horrified to hear that an interviewer, originally from Australia, called his client "partner". The banker said that it was unacceptable to address his customers in such a manner, and demanded that the interviewer be removed from further participation in his project. Oddly enough, this interviewer has conducted hundreds of surveys and never had a single complaint from consumers. He didn't try to hide his Australian accent and people took him for who he was with his usual manner of speaking. He never deviated from the survey design, only occasionally adding "partner" or "ok, partner" in response to a respondent's comments.

We can say that respondents are more willing to participate in the survey if the interviewer looks more natural and shows a sincere interest in what they tell him. However, caution is also needed here. It is unacceptable, for example, if the interviewer says “excellent” every time the respondent gives the highest rating on the proposed scale - thereby the interviewer, unwittingly, psychologically encourages the respondent to give only positive answers.

How to Maintain Interviewer Interest in a Tracking Project

Permanent control is not the only way maintain a high standard of interviewer performance. A highly qualified interviewer never forgets that he has conducted hundreds of surveys on a particular project, but for most participants this survey is something new in life. Therefore, it is very important that the interviewer shows the same interest in the words of the respondent both in the 50th and 100th time as in the very first time. In order to maintain interest in the tracking project, the agency conducts regular follow-up sessions with interviewers, where they are repeatedly reminded of the purpose of their work and its significance to both the client and the agency.

However, the most effective long-term motivation occurs when agency employees show a sincere interest in the progress of the project. This includes the active pursuit of feedback from interviewers (it’s good if interviewers see how the results they get immediately go to work). This also includes situations where a representative of the client company is so interested in the results of the work that he is ready to accompany the interviewer during the interview. This commitment is a powerful incentive to work, along with due respect for the skills and abilities of the interviewers and the contribution they make to the success of the project.

Of course, it is good practice to regularly improve the skills and abilities of interviewers on each specific project. This is especially important if, during the course of the survey, any changes are made to the questionnaire, the survey group or the goods and services that are the subject of the survey change.

Participation and interest of the client at the survey stage

Many companies take a special interest in the level at which their staff works with customers. Companies often invest heavily in training their salespeople, service personnel, and salespeople to deal with customers in a variety of situations. This is very typical for those companies where customer service is considered a strategic indicator. Surely, such a company will want to make sure that the interviewers that the agency attracts to carry out its project properly represent the company when meeting with the consumer. Many clients try to provide the agency with supporting materials - a detailed description of goods and services or a dictionary of the most common terms that consumers can use during a survey. Sometimes the company transfers to the agency Additional materials, such as internal videos of staff sessions or documents that prescribe which "voice tones" its employees should use when communicating with a customer.

If time permits, the client representative should attend briefing sessions with the interviewers. As a result, interviewers will not only receive a message from the company that it is truly committed to the success of the project, but will also be able to solve a variety of problems that arise in the process. It also avoids unnecessary delays in starting field work, which almost always requires additional clarification from the client.

Listening to customer satisfaction surveys

Clients often express a desire to attend at least a few surveys that are conducted for their project. There can be many reasons for this. To name just a few of them:

The desire to listen to several interviews on the newly created questionnaire to make sure that the person does not have any difficulties in understanding the questions, and their interpretation coincides with the intention of the questionnaire developers. If such listening is conducted at the beginning of the survey, it is considered as a test interview.
The desire to get an idea of the degree of interest and intensity of feelings of respondents in the subjects covered in the survey. Survey data will show the client the number of people answering each question, but will not convey the feelings that people show. Attending surveys gives an idea of the level of emotion that respondents show in relation to the topic under discussion.
Desire to hear how interviewers deal with phone calls, namely their objectivity, ability to manage the conversation and how responses are recorded. These are extremely important questions, so the client needs to make sure that the interviewers have the skills to not influence the answers of the respondents (which can lead to skewed results). There are many ways an interviewer can influence a respondent's responses, such as:

Questions a client asks when listening to an interview

If a client does decide to attend an audition, agency staff naturally expect them to start asking questions or sharing observations. The client can ask different questions, depending on the purpose of the visit, the depth of his knowledge in the field of market research, his position and responsibilities in the company. For example, when attending a trial ("pilot") session, the client may ask the following questions:

What questions did your interviewers have during the briefing and preparation sessions?
What do your interviewers think of the questionnaire?
What questions did your interviewers find difficult to ask, and what questions did you find difficult to answer?
Is it possible to see the contents of guidance materials for interviewers?
Why do you think respondents misunderstand the question X?
How long does one interview take?
What if? .. (Followed by various proposals for changing the wording in the questionnaire, the order of the questions, etc.).

If the client representative does not have much experience in market research, the range of questions expands, for example:

How do you find people to participate in this study?
How many refuse to participate in the survey?
How representative can the opinions received from the respondents be considered if I myself always refuse to participate in the survey under the pretext that I do not have time for this?
What level of training do your interviewers receive?
Are the answers we hear in this study much different from other surveys you have done?
Who are your other clients?
Can you provide us with the names and addresses of people who participated in the survey - we assure you that many of them will benefit from being contacted by our service department?

More experienced researchers may also ask about the level of feedback, sampling principles, etc.

Some issues may be related to the duties that the client representative performs in his organization. For example, employees of the marketing department will be interested in the statements of respondents about the "image" of the company, their attitude to the goods, services, advertising and competitors of the company. The customer service department and the call center's customer relations staff will be interested not only in respondents' views on service levels, but also in how interviewers connect and interact with survey participants. Operations personnel are generally interested in consumer attitudes towards the level of operation of the company's operating systems (or lack thereof).

Agency and its consumer

If you do not pay due attention to the client's visit, agency staff may underestimate the impact that even small details can have on him. A client's visit is an opportunity to show how important the agency attaches to working with its customers. On the eve of the visit, for example, it is desirable for the agency to receive a list of guests and their estimated time of arrival. This will enable the receptionist or interviewer (if the client intends to be present for the in-person interview) to properly meet and greet the company representatives. It is necessary to find out if the guests have any specific requirements (presence in a group of people with disabilities, ordering a taxi to meet at the station and on the way back to the station). Naturally, interviewers need to be prepared and instructed about the visit, decide organizational issues associated with the visit, such as lunch break, etc. Documentation that may be required during the visit should also be prepared in advance, such as confidentiality pledge forms, a paper copy of the questionnaire, a set of guidance materials for interviewers, and pens and paper for notes. If guests are to be met at the central office, it is necessary to take care of the proper preparation of the premises. If a client needs to go to the restroom and sees that the sanitary and hygienic standards are not observed there or the necessary supplies are missing, he may think that the agency does not care about its staff and comfortable working conditions.

These days, not all companies follow the dress code. It is now customary to dress more casually at work than it was 10 or 20 years ago. However, in the company high class as before, employees who are assigned to organize a client's reception, or those who can be introduced to him, should dress more formally. In most cases, this involves elegant business-style clothing. Most clients wear formal attire during the visit - a formal suit or corporate uniform, so agency employees should dress accordingly.

Clients often notice the smallest details, so a welcoming smile and a firm handshake at a meeting will already show the hospitality of the host. Agency staff should also pay attention to signals coming from the client, such as how client representatives communicate with each other. If their communication is informal, agency employees should also not behave formally. If client representatives ask questions, answer them truthfully and concisely. The client understands that agency staff may this moment not know all the answers, so he is usually happy to get the information he needs later.

When preparing to receive a client, one should not forget that the purpose of his visit is to hear how the interviewer works with the respondent. If the client listened to the interview at the call center, he should be invited to meet, if desired, with those interviewers whom he listened to. Usually the visitor refuses the offer, but if he has such a desire, then this can be seen as a great opportunity to show the interviewer how much the work he does is appreciated. Such a meeting will take no more than five minutes, but it will show the client the full transparency of the agency's working methods and his trust in his staff.

After the client has left

Immediately after the client's departure, it is recommended to thank the interviewers for their input and collect feedback immediately on what was done on the high level and what could be improved. If any of the interviewers have questions, they should arrange to meet with the team leader, either on the same day or as soon as possible. In addition, it is recommended to write a summary of the results of the visit, print it and place it in a special file. The summary should include a list of guests, a list of all interviewers who were involved in organizing the visit, a list of key issues raised and minutes of their discussion, and an agreed follow-up plan. Finally, the agency researcher should write a memo to comment on accomplishments or failures, and summarize "lessons learned" to help organize the work better next time.

The need to control the correct use of sampling

As the number of surveys increases, it is essential to monitor how the sample is distributed and used and to identify problems as early as possible. For example, very often customer satisfaction surveys involve interviewing customers who have recently been in contact with the company. The company provides the agency with a (randomly) selected sample for the survey. But the agency may not know, and the company may not understand that the person who created the sample could distort it in a special way.

For example, according to the instructions, you need to interview people who have recently entered into correspondence with the company. The company intends to find out how satisfied consumers are with how its employees respond to written inquiries or complaints. But if the company employee who created the sample was not told the purpose of the survey, he could draw a sample of consumers to whom the company recently sent a marketing letter, and this does not meet the objectives of the study. But if the interviewers and agency employees know their job well, the existence of a problem becomes obvious already at the first interviews. Unfortunately, in such a situation, it is necessary to postpone the start of work on the project until a new sample is received.

Quota control

One of the most challenging tasks for the agency, it is compliance with quotas (groups) that can be set for field work. As mentioned earlier, selection by quotas (groups) is one of the most common methods for selecting respondents when conducting surveys on the level of customer satisfaction. Even if the client company or agency has information about the market, the study can set special quotas that will ensure that the profile of the respondents matches the sample or consumer base. These quotas may be based on demographic characteristics such as age or gender, or product usage, or other characteristics. When the project is “in the field”, some specific quotas may be set for interviewers, and their task will be to select those people for the survey who meet the requirements of the quota.

Sample use control

Another problem an agency may face is the use of sampling. If a client company provides a sample of its consumer base for research, it usually agrees with the agency in advance on the number of names that need to be included in the list in order to achieve the required number of surveys. There is no magic formula for this, each time this issue is resolved depending on the specific situation. For example, an agency may agree with a client that the number of names on the list should be 10 times the number of interviews set. The client may request that interviewers not call certain categories of customers, such as people whose phone numbers are not listed in the directory, or those who participated in any market research within the last 12 months. The more criteria are set, the more names from the original list will be lost for research. If the client cannot a short time to provide the agency with an additional list of names, it is very important that interviewers try not to “miss out” a single potential respondent from the population they have left. This may mean, for example, that the same number will have to be dialed (if we are talking about telephone survey) at least 10 times within five days in order to obtain consent to conduct an interview, and only after that it will be possible to refuse a potential contact or replace it with another one. Modern technologies make it easier for the interviewer to organize the work with the quota he has to reach. At the end of each working day, he can send data to the head office in electronic format, which will allow the system to monitor the progress of the project. If the interviewer has reached the quota set for him, the computer system will lock the sample file and the interviewer will no longer be able to get any more names from it.

The need for trial (pilot) studies

Pilot study is a term that means testing a questionnaire before starting a full-scale survey. There are many reasons to conduct a pilot survey.

Ian Brace(2004: 164-65) indicates 15 important reasons for these tests:

it is necessary to check how questions sound to the ear (they may look great on paper, but when spoken aloud they sound grandiloquent or unnatural);
it is necessary to check whether the interviewers understand the questions correctly (if they do not understand the question, then there is very little chance that the respondents will understand it);
it is necessary to make sure that the respondents understand the questions correctly (the questions may contain professional vocabulary);
it is necessary to make sure that there are no questions in the questionnaire that have a double meaning or a different interpretation, as well as questions leading or containing an answer;
it is necessary to make sure that the respondents will be able to answer the question (that is, the questionnaire contains questions that the respondents are able to answer);
you need to make sure that the questionnaire gives a sufficient number of response codes (if you miss any of them, this will lead to an increase in the number of “other” responses that may require coding at the analysis stage);
it is necessary to check whether the questionnaire can keep the respondent's attention during the entire survey;
it is necessary to find out whether the directing instructions for the questionnaire are clear to interviewers and respondents;
you need to make sure that the interview process runs smoothly (that is, the questionnaire has no or a minimum number of abrupt transitions between apparently unrelated topics);
it is necessary to establish how long the survey will take (this is related to the cost of the entire project);
it is necessary to check the text of the questionnaire for errors;
you need to check how the instruction works;
it is necessary to check how the technology works in the field.

There is no need to conduct many pilot surveys. For a business survey, it is enough to conduct 5-10 interviews. Perhaps more than 30 interviews would be needed for a consumer survey that has complex guides or groups for different segments, each with separate questions.

In some companies, it is believed that it is enough to conduct a test survey among colleagues. Of course, this will help save money and identify a number of problems, but will not replace the pilot study conducted by real interviewers among real respondents.

Pilot interviews with respondents

There are several ways to conduct pilot interviews with respondents. Thus, the researcher may accompany the interviewer who pre-screens the questionnaire in a face-to-face survey. At the end of the survey, the researcher may ask the respondent if there were any questions that were difficult to understand or difficult to answer. The researcher may also ask the respondent to provide additional information about what underlies their unexpected responses to the questionnaire. This may indicate that the question is not structured correctly, or that the survey should be entered additional element. If the interview is conducted by phone, you can use the same approach and ask questions at the end of the interview.

It is not always possible to accompany the interviewer or listen to pilot surveys. If this is not possible, feedback from the interviewers should be obtained as soon as possible after the completion of the pilot survey. In these cases, it is recommended to ask the interviewers to take notes on all the details of the pilot interviews and then review them with the interviewers. Often the smallest details can make a difference. For example, the questionnaire asks the consumer about the reason for his last visit to the company's office and suggests choosing only one item for the answer. But the respondent may answer that he visited the company in order to resolve a number of issues or conduct several transactions, because he was accustomed to "accumulating" cases in order to do them all at once. If the interviewer can only code one response, this will skew the information collected. Moreover, if this question is further used to analyze the drivers of customer satisfaction, the distortion of information will only get worse.

Some companies prefer, instead of conducting a pilot study, to consider as such the first interviews on initial stage implementation of the project (i.e. to analyze the first series of planned surveys). Companies believe that this saves time and, if such an analysis does not reveal any errors, it saves money on conducting a separate pilot survey. Such savings can result in a variety of problems. First, the agency must suspend surveys in order to make possible changes to the questionnaire or to the course of the study. The interviewers will then have to be re-instructed before proceeding with the project. If it is a large-scale study, it involves a large number of interviewers, and this stop-and-go approach is not only undesirable in itself, it can negatively affect the enthusiasm and interest of the interviewers working on the project. It also disrupts the work schedule of the researchers, who are forced to enter into negotiations with the client about a break in the progress of the project and agree on a new schedule. This creates an inconvenience for analysts and programmers who have to interrupt work on other projects in order to make changes to this one. Not only that, the client may also have to change the internal schedule due to these delays.

Solving serious questions that arise during the interview

Sometimes, when conducting customer satisfaction surveys, respondents raise very serious questions. For example, they may provide examples of fraud or other potentially illegal behavior. Respondents may express disappointment that their concerns or complaints are not addressed, or talk about a situation that led to stress or financial loss. How should the interviewer, and with him the agency, deal with such a problem? After all, the introductory part of the questionnaire contains a reminder to respondents that all information they provide will be considered confidential.

The actions an agency can take depend on the specific situation. In the UK, for example, a shop code of conduct does not take precedence over the law, and the Market Research Society of Great Britain has set up an advisory service for its members to help in such situations.

However, if the agency is able to resolve the issue raised by the respondent, then with the written consent of the respondent, the agency may forward the details of the problem to the client. In this case, the respondent must know exactly what information will be shared with the company, and their written consent must be obtained before the information is shared. Respondents should also be told (if possible) to whom in the company the information is being shared. No other data collected from the respondent during the survey should be sent to the client (see next section).

The Importance of Client Confidentiality

Market research depends on voluntary consent citizens and employees of consumer organizations to allocate time to participate in the survey. In order to maintain their credibility, it is important that all market research be conducted transparently—openly and honestly, in a way that is objective, non-damaging, and non-intrusive. In addition, the market researcher must respect the confidentiality of the respondents and not use the data obtained so as not to harm or embarrass the respondents. The researcher must also ensure that he does not misrepresent the data.

Without respondents, there would be no market research - this is an indisputable fact. Therefore, in most countries where there is a developed market research industry, there is also a professional body with its own code of conduct.

How to win the trust of respondents

One of the most common questions a researcher is asked by people invited to participate in a study is: who is it for? Potential respondents usually want to make sure that they are not trying to impose some product (sell under the guise of research). In addition, they want to know who the sponsor is because they care if they are heard by the company.

In some cases, especially when it comes to customer satisfaction surveys, it is necessary to name the customer of the survey at the very beginning. This is inevitable, for example, if the interviewer has to ask about the last transactions that were performed for the respondent in the organization, and the organization has provided a corresponding list of customers. In other cases, clients prefer to be identified at the end of the survey. However, it is not always possible for the interviewer to tell the respondents who commissioned the survey, as this may influence the responses of the respondents and thus distort the results of the survey. Therefore, it is important that interviewers win the trust of respondents by being open and frank, to the extent feasible.

Sometimes the respondent, knowing the name of the sponsoring company, forgets that he is actually talking to an employee of the research agency, begins to go into details about the goods or services (which, in principle, can give a lot useful information), but his comments do not fit the given context. It is in such situations that the skill of the interviewer is manifested, since it is important for him to appear before the respondent as a plenipotentiary ambassador both from the agency and from the client company.

Final comment

The director of a market research company once said, “It’s not recommended for a researcher to go out into the street and watch a survey being conducted, it’s like visiting a front-line casualty evacuation point during a war is not recommended for a general!” However, this chapter shows the need to know and control everything that actually happens during the survey.

In a report on the first small clinical trial in the field, researchers at Johns Hopkins University's Kimmel Cancer Center say they have safely used immune cells grown from patients' own bone marrow to treat multiple myeloma, a cancer of the white blood cells.

This figure shows the infiltrating Bone marrow lymphocytes in cell culture

The results of the study involving a specific type of tumor-targeting T cell known as marrow-infiltrating lymphocytes (MILs) have been described in the journal Science Translational Medicine.

“What we learned in this small study is that a large number of activated MILs can selectively target and kill myeloma cells,” said Johns Hopkins University immunologist Ivan Borrello, MD, who led the clinical trial.

He explained that MILs are foot soldiers of the immune system and attack foreign cells such as bacteria or viruses. But in their normal state, they are inactive and too few in number to have a noticeable effect on cancer.

Previous laboratory studies by Dr. Borrello and colleagues have shown that activated MILs can selectively target and kill myeloma cells taken from patients and grown in the laboratory.

For clinical trials, the research team recruited 25 patients with newly diagnosed or recurrent multiple myeloma, although 3 of these patients relapsed before they could receive MILs treatment.

The scientists isolated MILs from each patient's bone marrow, grew them in the lab to increase their numbers, activated them with microscopic beads coated with immune-activating antibodies, and injected each of the 22 patients intravenously with their own cells. Three days before the introduction of improved MILs, patients received high doses chemotherapy and stem cell transplantation are standard treatments for multiple myeloma.

One year after treatment with MILs, 13 of 22 patients had at least a partial response to therapy, meaning that their cancer had been reduced by at least 50%.

Seven patients experienced volume reduction. tumor cells by at least 90% after an average of 25.1 months without cancer progression. The remaining 15 patients had, on average, 11.8 months of disease progression-free time after treatment with MILs. None of the participants had serious side effects from the treatment. Overall survival was 31.5 months for those with less than 90% disease reduction. The mean follow-up time is currently over 6 years.

Dr. Borrello noted that several US cancer centers have conducted a similar experimental treatment known as adoptive T-cell therapy, but said the Johns Hopkins team believe they were the only ones using MILs. He said other types of tumor-infiltrating cells could be used, but they tend to be less numerous in a patient's tumor and may not grow well outside the body.

In tumors like melanoma, only half of the patients have T cells in their tumor that can be isolated, and about half of which can be grown. “Typically, immune cells from solid tumors, called tumor-infiltrating cells, can be isolated and grown in as few as 25% of patients who could potentially be treated. But in our clinical trial, we were able to isolate and grow MILs in all 22 patients,” said Kimberly Nulan, Ph.D., a research assistant at the Johns Hopkins University School of Medicine.

Nulan says the small study helped her and her colleagues learn more about which patients might benefit from MILs treatment. For example, they were able to determine how many of the MILs grown in the lab were specifically targeting the patient's tumor and whether they continued to target the tumor after injection.

In addition, the researchers found that patients whose bone marrow before treatment contained higher amounts of certain immune cells, known as central memory cells, also had a better response to treatment with MILs. Patients who started treatment with signs of an overactive immune response did not respond well to treatment.

Dr. Noonan says the research team has used the data to guide two other ongoing clinical trials of MILs. These studies are trying to expand the antitumor response and tumor specificity by combining transplantation of MILs with a cancer vaccine called GVAX, developed at Johns Hopkins University, and the myeloma drug lenalidomide, which stimulates T-cell responses.

The researchers also say the trial has also shed light on new ways to grow MILs. “In most of these studies, you see that the more cells you get, the better the response will be in patients. Learning how to improve cell culture can improve treatment,” Dr. Noonan said.

Kimmel Cancer Center scientists have also developed MILs to treat solid tumors such as lung, esophageal, and stomach cancers, as well as childhood neuroblastoma and Ewing's sarcoma.

Experts report that more than 20,000 cases of multiple myeloma occur each year in the US, and more than 10,000 people die from the disease. It is the second most common blood cancer.

The term "pilot study" in the literature is used as: 1) a synonym for exploratory (reconnaissance) research; 2) a synonym for pilot study; 3) a collective concept that includes both its varieties (subspecies) exploratory (reconnaissance) and pilot studies. We will consider that search (reconnaissance) and pilot studies are the two main subtypes of pilot research.

Search (reconnaissance) research is carried out with the aim of clarifying the problem, more correct formulation of problems and putting forward well-founded hypotheses. Thus, at its core, it is a conceptual study. It is especially desirable to conduct it if there is no literature on the subject of interest to you or it is not enough.

Intelligence research is the simplest type of sociological analysis: the range of tasks is limited, the number of respondents is small, the program and tools are extremely simplified; the data is not representative. The scientist receives only tentative information about the object of study for a general orientation in the problem. It is used for little studied or not studied problems at all. Its slogan is about, cheap and fast.

Exploratory research can be conducted by the following methods: interviews (preferably informal) with potential respondents; observation; focus group on the central research problem; a survey of experts - specialists or just people related to the problem area that interests you; study of documents, statistical data containing the necessary information on the tasks and hypotheses put forward.

Express surveys are a variant of intelligence research. They are also called operational surveys. . They are used by numerous polling firms - from VTsIOM to ROMIR. True, they conduct surveys, as a rule, are very competent, but they do not set deep scientific tasks for the development of fundamental science. Momentary, but very necessary utilitarian tasks for society, a department or a private customer are being solved: how the people treat the president, the ban on abortion, the war in Chechnya, the arrival of Bush, the terrorist attacks of September 11, 2001. Thus, they get a fresh, short-lived (its life span is calculated in days , weeks, less often months), but information that is very necessary at the moment.

An exploratory study is also often referred to as a pilot study. Although it is more correct to consider it an independent type of sociological research. Reconnaissance and pilot studies are similar in two respects:

♦ Target - obtain approximate data on a particular phenomenon or test the application of the methodology to larger-scale studies.

♦ Object - to conduct both studies, an insignificant set of objects is needed, therefore they are carried out in a limited time frame.

But unlike reconnaissance, a pilot study is designed to work out technical procedures and techniques, most often to test a questionnaire. According to J. Mannheim and R. Rich, a preliminary check of the survey instrument is as important for the success of the study as a test drive is for a successful purchase of a used car. It helps to detect problems that can only fully manifest themselves in the field.

In sociology, aerobatics is carried out before the main study and acts as a way to test the validity of hypotheses and tasks, as well as the professional level and methodological sophistication of the tools. Piloting helps to evaluate the correctness of the sampling model and make appropriate corrections to it, if necessary; clarify some characteristics of the object and subject of the study, justify the financial costs and timing of the main study. Piloting is also useful for training a group of interviewers (questionnaires).

Piloting is useful for: a) testing the instrumentation in which the researcher is quite confident, b) improving the instrumentation in a situation where the research subject is less known to the researcher. In the first case, the toolkit passes a preliminary test in its final version. In the second, the researcher may want to experiment with different versions (layouts) of the tool to find out which one is more convenient to work with.