Concepts Flashcards
Test of sphericity
when conducting ANOVAs with repeated measures (within subject factors), your factors become vulnerable to violating the assumption of sphericity. this assumption assumes a condition where the variances of the differences between all combinations of related groups are not equal. this assumption is violated when these variances become unequal leading to an increased risk of type 1 error (ADD EXAMPLE HERE). to correct this , this test estimates the degree to which sphericity has been violated allowing one to apply a correction factor to the degrees of freedom of the f distribution
simple regression
simple regression can be used to model the relationship between two continuous (or ordinal) variables. this can help predict the value of an output variable/response based on the value of the input/predictor variable. both variables should be quantitative with one being dependent and one being independent. For example, y=a+bx, used to make prediction with x value based on an unknown y value/
methods section
in the methods section of a research report, a researcher will describe the steps they included to conduct their experiment. this includes key information such as who the participants are (including age, sex and how they were sampled). this also includes describing which measures were used to collect data e.g. observation, design of task and resources used. a methods section should contain enough detail that somebody could use it to replicate the experiment although it usually is confined to a word limit.
discourse analysis
discourse analysis describes a set of methods used for studying the contents of people’s speech and interaction in conversations. this method could be used, for example, to determine if younger people use more filler words in their speech than older people. or how tone varies with age. some issue with this method is that it can be time consuming and not all information can be analyzed. it can also be directional leading to interpretational data rather than testing the hypothesis
floor effects
the situation in which a large proportion of participants perform very poorly on a task or other evaluative measure, thus skewing the distribution of scores and making it impossible to differentiate among the many individuals at that low level.
independent samples t test
an independent samples t test is used when two independent groups of participants are tested to examine the effects of a single IV. For example, we might test the effects of alcohol intake on reaction time by administering two different levels of alcohol to two different groups (low alcoholic and high alcohol) and collecting reaction time data. the results of the analysis would tell you if the two level of the IV had a significant effect (whether alcohol levels increased or decreased reaction time significantly)
post hoc tests
a significant f value tells us that there is a difference between means somewhere but not where the difference lies, or which groups/levels differ from each other. Post-hoc tests are after the fact comparisons and thus should be corrected to avoid inflating alpha. the test will give a p-value for the difference which can then be interpreted. For example, we might test the effects alcohol intake on reaction time by administering three levels of alcohol (none, low, high). a significant main effect would indicate that there is a difference between the three means, and the results of the post-hoc test would tell you which means were significantly different. For example, only high alcohol might differ from both low and none but low and none do not differ from each other.
one way analysis of variance
a one way ANOVA is used when one independent variable is manipulated with 3 or more levels. It can be between subjects (different participants in each group) or repeated measures (with the same participants measured at three time e.g. before, during and after a treatment) The results of the analysis tell you if the IV had a significant main effect (for example whether responses differed over time of treatment) The ANOVA separates out the variance in data into that explained by the IV and the unexplained or error variance. The F value is simply the ratio of these two things: explained v unexplained.
informed consent
informed consent implies that participants should know all relevant information about an experiment before agreeing to take part. the BPS guidelines argue that informed consent is a crucial aspect of ethical experimental design. However, they also acknowledge that mild deception is sometimes necessary to permit investigation of certain kinds of psychological phenomena. In these cases, participants should be fully debriefed at the end of the study. Further, deception is not considered appropriate if participant are likely to object to the study once they are made aware of its true purpose. Informed consent does not mean telling participants absolutely everything about a study in advance, but it does imply that nothing that might influence decisions to participate should be withheld from them.
discussion section
the function of a discussion section is to interpret results in relation to the experimental hypothesis and to discuss what the findings mean. a discussion section should start with a short summary of the study results and get broader throughout ending with a section on the broad implications of these results for the field. the discussion should talk about the results in the context of literature, what new information or interpretation do the results provide. It should also discuss the possible limitations of your study, but should comment on how they might have influenced your results and what you should do about it next time. The discussion can also propose future studies that would answer questions arising from results. Finally, it should end with a summary of the main conclusions.
Grounded theory
Grounded theory facilitates discovery and helps develop new ideas by approaching phenomena afresh. It involves intensive examination and re-examination of cases and extraction of structure. Provisional models from analysis are reapplied to existing and new data, therefore it is considered a self-correcting method. provisional conclusions are communicated to participants who comment on their plausibility thus providing respondent validation. Data is first collected from theoretically interesting cases, before it is coded, the idea of coding is to constantly compare similarities and differences between identified concepts and themes. The core analysis occurs using memo, writing, definitions and integrating categories. These are then applied to data collection or coding and the process continues. The outcomes of grounded theory are key concepts, definitions, memos and relationships/models.
cluster analysis
cluster analysis is a method used for finding data structure, it looks for categories withing data. you would use cluster analysis to classify participants or objects into categories based on similarity. For example, if you were interested in confidence you might want to group participants into a small number of categories according to their self-ratings and how other people rate them to get accurate confidence categories. The cluster analysis will enable the researcher to create subgroups of participants. These subgroups could then be used to look at the link between confidence and sport performance in more detail.
Moderated effects
A moderator is a variable that modifies (moderates) the relationship between a predictor and an outcome. Moderated effects therefore implies that predictor’s effect on the outcome depends on the value of another predictor (moderator). Assessing moderation helps to answer “when” questions. For example, a researcher may want to know if the relationship between stress on well-being differs according to level of social-support. If there is a moderated effect, it tells the researcher that the relationship does differ depending on level of social support. For example, while there is a positive relationship between stress and well-being when social support is low, there is no relationship between stress and well-being when social support is high.
spurious effect
spurious effects are where there is a significant relationship between two variables that is explained by a third unmeasured variable that is related to both of the other measured variables. This means that the two measured variables are not actually related to each other. For example a researcher may find a relationship between sunburn and drowning, which are actually not related but both are related to temperature. Spurious effects can be tested for using hierarchal multiple regression, if the significant relationship becomes ns once the third variable is entered into the regression equation, then the original relationship was spurious.
event-contingent recording
event contingent recording is a time-sampling method. this method measures variables on many more occasions, usually separated by relatively short intervals, allowing a closer focus on how responses unfold over time. Participants provide data whenever a pre-specified event occurs, permitting researches to focus investigation on particular situations (but not to sample a wide variety of experiences). Anticipation of events leads to the problem of reactivity, where anticipation becomes a significant event in its own right.
