randomization
Assignment by chance in studies
the process of putting people or things into study groups by chance, not by choice
randomization of [participants/patients/samples] to [groups/treatments/conditions]
- Randomization assigned half the patients to the new drug and half to standard treatment.
- Good randomization helps reduce bias between the treatment and control groups.
- The study was criticized because it did not clearly describe its randomization procedure.
- Randomization put the patients into two groups by chance.
- The trial used randomization to make the results fairer.
Making order or choice random
mixing or choosing things by chance so they do not always appear the same way
randomization of [items/questions/order/results]
- The app uses randomization so the quiz questions do not appear in the same order every time.
- Playlist randomization kept the music from feeling repetitive.
- The game includes randomization of rewards to make each round less predictable.
- The test uses randomization to mix the questions.
- Randomization makes each game a little different.
Adding unpredictability in software
using chance in software so that locations, numbers, or results are harder to predict
randomization of [addresses/identifiers/values/output]
- The browser adds randomization to certain identifiers to make tracking harder.
- Memory randomization makes it more difficult for attackers to predict where code is stored.
- The game uses map randomization so each new world is different.
- The software uses randomization for extra security.
- Randomization helps make the game maps different each time.
- Base: randomization
- Plural: randomizations
address space layout randomization
1Randomized memory locations
a security feature that changes where program parts are placed in memory so attackers cannot easily guess their location
- Modern operating systems use address space layout randomization to make memory exploits more difficult.
- The vulnerability was easier to exploit when address space layout randomization was disabled.
- Address space layout randomization is one layer of defense, not a complete security solution.
block randomization
1Balanced assignment in blocks
a study method that randomizes people in small sets so the groups stay about the same size
- The researchers used block randomization to keep the treatment and control groups balanced during recruitment.
- Block randomization with varying block sizes helped protect the assignment sequence.
- Each clinic used its own block randomization list.
cluster randomization
1Random assignment of groups
randomizing whole groups instead of single people
- The researchers used cluster randomization, assigning entire schools to the new program or the usual curriculum.
- Cluster randomization reduced contamination between students in the same classroom.
- The public health study used village-level cluster randomization.
mendelian randomization
1Causal study using genetics
a research method that uses genetic differences people are born with to study possible causes of disease
- The paper used Mendelian randomization to examine whether higher BMI increases the risk of asthma.
- Mendelian randomization can strengthen causal evidence when a normal randomized trial is not practical.
- The Mendelian randomization analysis found little support for a causal effect.
randomization ratio
1Planned assignment proportion
the planned numbers of people put into each study group by chance
- The trial used a 2:1 randomization ratio, so more patients received the experimental treatment.
- A 1:1 randomization ratio gives the same planned number of participants in each group.
- The protocol changed the randomization ratio after the dose-finding stage.
randomization test
1Significance test by reshuffling
a statistics method that checks whether a result could happen by chance by repeatedly mixing the data in random ways
- The analyst ran a randomization test by shuffling the treatment labels 10,000 times.
- A randomization test was useful because the sample was small and the usual assumptions were uncertain.
- The randomization test showed that the observed difference was unlikely under the null hypothesis.
simple randomization
1Unrestricted random assignment
putting study participants into groups by pure chance, like using a coin toss or random number
- The trial used simple randomization because enrollment was large enough to keep the groups fairly balanced.
- With simple randomization, small studies can end up with unequal numbers in each group.
- The software performed simple randomization for each participant at the time of enrollment.
stratified randomization
1Randomization within subgroups
randomizing people separately inside groups with the same important feature, such as age or clinic
- The trial used stratified randomization by age and disease severity.
- Stratified randomization helped ensure that each treatment group had similar numbers of high-risk patients.
- The protocol specified stratified randomization within each study site.