More on experiments, confounding and obscuring variables

36 important questions on More on experiments, confounding and obscuring variables

How do we call a designing mistake where the researcher doesn't design the independent variable well enough and therefor anther variable happens to vary systematically along with the intended variable therefor being an alternative explanation for the results?

a design confound.

How do we call mistakes in the results that occur due to the fact that the different groups of participants differ systematically from each other?

selection effects

How do we call the principle where a change in behavior changes naturally and spontaneously over time without any influence?

the maturation threat.
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

How do we call spontaneous improvement with a patient with a disorder without a clear cause for this?

spontaneous remission.

How can you tackle the maturation threat?

by designing a study with a control group.

How do we call external factors that influences the dependent measure, that affect almost all subjects in a study at the same time as the treatment which could cause confounding. things like winterdepression and springhappiness in a study on depression treatment.

history effects.

How can you tackle history effects an why so?

by using a comparison group, if history effects occur the control group should be influenced by this too.

How do we call the principle where the group average on a score is reduced in the posttest, compared to the pretest where this is caused by extreme scorers in the pretest condition simply scoring less extreme in the posttest condition, even without treatment in between?

regression to the mean.

How come with regression to the mean scorers who did score more extreme scores didn't do so in the posttest?

a lot of factors can influence a mood or a level of depression or something on a day, because of this someone could simply have a really bad or really good day. it is unlikely that these factors are similar on the day of posttest so the score will be less extreme then.

How can you detect regression in a line graph of two groups?

if the treatment group and the control group don't start of at the same value. one group will be more extreme, decrease in level can be accounted to regression but extreme decrease in the therapy group can not be accounted just to regression the treatment must have worked too.

How do we call the principle that can change the mean in a pretest-posttest situation because a specific type of participants are unavailable to show up for the posttest. because of this principle the mean can drop if the highest scoring participants don't show up for the posttest.

attrition threat.

How can you tackle attrition threat?

if a participant drops out of the posttest, simply don't count their score in the pretest either.

How do we call the principle where there's a change in a participants result as a result of taking the test more than once?

testing threat.

What's the difference between testing threat and practice effects?

with practice effects the subject only gets better at it so the score gets better, testing threat is a general term for influence of repeated test taking which includes getting a higher score due to practice effects as well as getting lower score due to getting bored.

How do you tackle testing threat?

  • use posttest only design
  • use alternative forms of the test for pre and posttest
  • comparison group.

How do you call a principle where a measuring instrument changes over time between pre and posttest, like an observer's judgement towards behavior?

instrumentation threat.

How do you tackle instrumentation threat?

  • use a posttest design only
  • use clear coding manuals for the observers.

How do you call an external factor that only effects one level of the independent variable?

a selection-history threat.

How do you call the principle where consistently for only one level of the independent variable people don't show up for the posttest instead of for both groups?

selection-attrition threat.

How do we call the principle where the expectation of the observer influence their interpretation of the results?

observer bias.

How do we call the principle where participants guess what the study is about and change their behavior in the expected direction of the study?

demand characteristics.

How do you call the principle of people actually improving after getting what they thought was a legit treatment when it actually wasn't?

placebo effect.

How do you call a study design where you use a placebo group, and a treatment group but neither the observer as the subjects know who is in which group?

a double-blind placebo control study.

How do we call when the independent variable does not make a difference in the dependent variable?

a null effect.

What could be possible reasons or a null effect?

  • weak manipulations
  • insensitive measures
  • ceiling and floor effects

Which reason for a null effect is described; the differences between the levels of the independent variable aren't big enough to see a difference in dependent variable. or the levels of independent variables aren't extreme enough in general to influence the dependent variable.

weak manipulations.

Which reason for a null effect is described; when the instruments can't measure precise enough to detect the small differences in the results of the experiment, which makes it look like there isn't a difference when there actually is?

insensitive measures.

How do we call the principle that occurs when all the scores on a test are squeezed together at the maximum side of the possible scores?

a ceiling effect.

How do we call the principle that occurs when all the scores on a test are clustered on the low end of possible scores?

a floor effect.

How do we call the principle of too much unsystematic variability within each group in an experiment?

noise.

How do we call a human, or instrument factor or influence that can inflate of deflate the score of a subject on a measurement?

a measurement error.

What influences variability within groups?

  • measurement errors
  • individual differences
  • situation noise

How do we call external factors that cause variability within groups?

situation noise.

How can situation noise be tackled?

by carefully controlling the environment an experiment takes place in.

How do you call the probability that a study finds an effect, if there's indeed an effect in the population?

statistical power.

How can you increase the power of an experiment?

  • increase the manipulations of the independent variable
  • use bigger samples to decrease the standard deviation
  • decrease variety within groups to decrease standard deviation
  • decrease level of confidence

The question on the page originate from the summary of the following study material:

  • A unique study and practice tool
  • Never study anything twice again
  • Get the grades you hope for
  • 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
Trustpilot Logo