C - Grant et al Flashcards
Aim
aimed to show that environmental context can have a more positive effect on performance in a meaningful memory test when the test takes place in the same environment in which the to-be-remembered material was originally studied (the matching condition) than when the test occurs in a different environment (mismatching condition
Background
- Research has shown that context-dependence may play an important role in numerous situations, such as memory for studied material.
- Grant et al were interested in determining whether environmental context-dependency effects would be found with the type of material and the type of tests typically encountered in school.
- Their focus is more on study conditions than on differences in classroom testing conditions because they hold that students have more control over their study environments than over their test environments.
- Observations had shown them that many high school and college students study material in environments very different from those in which they are tested: study environments often include background noise from either family, friends or television, while test environments are typically quieter. Therefore, if context-dependency occurs with meaningful course material, students’ study habits could be harming their test performance.
Method
- This was a laboratory experiment using an independent measures design.
- The independent variables (IVs) were:
(i) whether the participant read the two page article under silent or noisy conditions
(ii) whether the participant was tested under matching or mismatching conditions - The first IV – study context (silent versus noisy) and the second IV – test context (silent versus noisy) were manipulated in a between-subjects factorial design, producing four conditions.
- The dependent variable (DV) was the participant’s performance on (a) a short-answer recall test and (b) a multiple-choice recall test.
Sample
- Eight members of a psychology laboratory class served as experimenters. Each experimenter recruited five acquaintances to serve as participants.
- There were 39 participants, ranging in age from 17 to 56 years (M = 23.4, SD = 5.9), 17 were female, 23 were male. (1 participant’s results were omitted from the analyses.)
Procedure
- Each experimenter ran one participant for each of the four conditions and an additional participant for one of the conditions as assigned by the instructor. Experimenters randomly assigned their participants to their five conditions.
- Stimuli
(a) Each experimenter provided his/her own cassette player and headphones. The eight cassettes were exact copies made from a master tape of background noise recorded during lunchtime in a university cafeteria. The background noise consisted of occasional distinct words/phrases embedded within a general conversational hum that was intermixed with the sounds produced by movement of chairs and dishes. The tape was played at a moderately loud level.
(b) A two-page, three-columned article on psychoimmunology (Hales,1984) was selected as the to-be-studied material.
(c) 16 multiple-choice questions, each consisting of a stem and four alternatives were generated, all of which tested memory for points stated in the text. 10 short-answer questions were derived from those multiple-choice stems that could easily be restated to produce a question that could be answered unambiguously by a single word or phrase. The order of the questions on each test followed the order in which the tested points were made in the text. The short-answer test was always administered first to ensure that recall of information from the article was being tested and not recall of information from the multiple- choice test. - Instructions, describing the experiment as a class project and stating that participation was voluntary, were read aloud.
- Participants were asked to read the given article once, as if they were reading it for a class assignment. They were allowed to highlight and underline as they read.
- Participants were informed that their comprehension would be tested with both a short-answer test and a multiple-choice test.
- All participants wore headphones while they read. Those in the silent condition were told they would not hear anything over the headphones whilst those in the noisy condition were told they would hear moderately loud background noise, but that they should ignore it.
- Reading times were recorded by the experimenters.
- A break of approximately two minutes between the end of the study phase and the beginning of the test phase was incorporated to minimise recall from short-term memory.
- The short-answer test was given, followed by the multiple-choice test.
- Participants were tested in either silent or noisy conditions and were informed of the condition before testing. Regardless of testing condition, all participants wore headphones.
- At the end of the testing phase participants were debriefed concerning the purpose of the experiment.
- The entire procedure lasted about 30 minutes.
Results
- Results suggest participants in all groups spent roughly equal amounts of time studying the material. Therefore reading time was used as a co-variable in the analysis of test performance.
- There was a reliable Study Condition x Test Condition interaction for both the short-answer test and the multiple-choice test. A planned contrast comparing performance in the matching conditions (silent study/silent test and noisy study/noisy test) to performance in the mismatching conditions (silent study/noisy test and noisy study/silent test) was reliable showing that studying and testing in the same environment produced better results.
- There was no overall effect of noise on performance.
Conclusions
- There are context-dependency effects for newly learned meaningful material regardless of whether a short-answer test or a multiple-choice test is used to assess learning.
- Studying and testing in the same environment leads to enhanced performance.
- Students are likely to perform better in exams if they study for them with a minimum of background noise because, although there was no overall effect of noise on performance, the fact that there was evidence for context-dependency suggests they are better off studying without background noise as it will not be present during actual testing.
Method Evaluation
P: Conducted in a controlled environment – lowered EV’s
E: This allows us to establish cause and effect (IV had affected the DV) = high internal validity
E: all P’s listened to the same background noise from a master tape, they all read the same article and completed the same short answer and multiple choice questions, they all had 2 min break in between finishing reading and taking the test and they all completed it in the same order.
P: However it was artificial situation = therefore demand characteristics could have impacted the study
E: P’s could have guessed the aim and then acted to benefit the study/or throw it
P’s may have realized that the study was about learning with/without noise as they were all wearing headphones. The task was artificial in some ways as we don’t necessarily learn and then are tested straight away
Data
P: Quantitative data
E: numerical form so easy to compare
Short answer questions and multiple choice answer questions were recorded so easy comparisons could be made between the conditions, we know that 14.3 mean recall was achieved in the matching conditions for the MCQ and only a mean on 12.7 in the mismatching
P: Not qualitative
E: There was no opportunity to explain why
P’s didn’t have the chance to say why they think they recalled as well/not so well. May have been due to a lack of concentration rather than whether learning and recalling was matched or not.
Ethics
the participants provided informed consent and voluntarily participated in the study of their own free will. Further, the participants were protected from harm and debriefed at the end of the study. Therefore the study was highly ethical.
Reliability
This study could be considered as high in reliability due to lots of standardisation put in place.
This means that the researcher ensures thatallparticipantshave the same experience. This increases thereplicability of thestudy and the likelihood that if repeated by another the same results would be found.
This study could be considered as low in reliability as not all elements of the study were standardised.
This means that not all participants had the same experienceand therefore makes itharder foranother researcher to repeat the study in exactly the same way, this reduces the chance that they would find the same results.
There were 8 experimenters, each had their own cassette player. This equipment could havediffered in its quality of sound (the uni café background noise), this may have changed the results as some P’s may have had better or worse results because of this.
Validity
P: High internal validity
E: This means that there is a high chance that only the IV is affecting the DV as there was a high level of control.
E: For example there was a 2 minute break between the study period and testing period to ensure that recall was not from short term memory but due to the matching or mismatching conditions.
P: Low ecological validity
E: May not be able to generalise these findings to everyday study habits in real life.
E: Students knew they were taking part in an experiment and therefore they may not have demonstrated their true study habits. They may have tried really hard in the experiment to please the researcher (demand characteristics) or they may not have been trying hard because they knew it was only an experiment that would have no impact on their actual studies.
Sampling Bias
This sample may be seen as a representative sample
This is because it included P’s who were both males and females. This increases thelikelihood that results can be applied tothetarget population.
Grant sample had similar amounts of males and females (17 f, 23m) read the article ofPsychoimmunology and tested via short answer and multiple-choice questions.
However, he gathered the participants using anopportunity sample
People who wereavailable at the time of the study were asked to take part. This often creates unrepresentative samples as these participants often share similar characteristics and nota good cross-sectionfound in the target population and therefore results cannot be applied toa wider group.
The sample in the matching and mismatching conditions were the experimenters acquittances (not necessarily students but more than likely) so this could led to a bias sample as the results would not representwhat would be found ifthe researcher studied non-students.
Ethnocentrism
This is because the sample only contained people from oneculture- this means that the results cannot be generalised (applied)to people of other cultural backgrounds.
In Grants study all the participants are likely to be living in America. A country with a western culture that may focus more attention and importance on learning in this way, studying written material and completingwritten assessments in a classroom environment. This study may then not be relevant to other culture wherecontent-dependency may still be important but in a different way.
