The aim of this study is to find out how being deprived of a normal upbringing effects feral behaviour.
The alternative hypothesis is that being deprived of a normal upbringing causes feral behaviour.
The null hypothesis is that having been deprived of a normal upbringing will have no effect on how feral a child’s behaviour is.
Scenario Three
A market research company wanted to gauge the attitudes towards a new product before it comes onto the market.
The method used for this study is a questionnaire. A questionnaire is a non-experimental method. A questionnaire is a series of questions that are given to a participant on a sheet, for the participant to fill out themselves. Questionnaires may consist of closed questions, which have specific limited answers, or open questions where the participant can submit any answer of any length and detail. This method was chosen because it is a quick and cheap, but efficient method of measuring people’s attitudes. It is also a method that can be used to test a large sample, over a large geographical area, which is essential if a product is to be marketed nationwide.
The main advantage of using a questionnaire is that it is a standardised procedure, where there is similar treatment of all respondents. There is also less bias from interpersonal factors. There is greater reliability and consistency of data analysis. The results can be tested for reliability. The exact same questionnaire can be repeated, on both previous participants and a new sample. The answers to closed questions have the advantage of being quantifiable. This means that their answer can be scored to be used in comparison to other participant’s answers.
The disadvantage here is that if only limited answers are given, then the results obtained have no specific detail to the participant. Open ended questionnaires allow the participant to give a full, detailed description, however these results are difficult to quantify and comparisons cannot be made. The most common disadvantage of a questionnaire is that the misinterpretation of questions may produce distorted answers.
The independent variable in this study is the product. The dependant variable is the participant’s attitude towards the product. Their attitudes can be measured by asking questions about the product, where only closed answers can be given. Answers can either be positive or negative and the attitude of the participant can be measured by recording how many positive answers and how many negative answers are given.
The aim of this study is find out what attitudes people have towards a product, before it is marketed.
The alternative hypothesis is that participants will have a positive attitude towards the product.
The null hypothesis is that participants will be indifferent, and have neither a positive nor a negative attitude towards the product.
Scenario Four
A psychologist was interested in studying the effects of bereavement of a first degree relative. This obviously involves covering sensitive issues.
The method used in this scenario is an interview, specifically a semi-structured interview. An interview is a non-experimental method. An interview is where a participant discusses there thought and feelings about a particular subject with the interviewer. This is often recorded. An unstructured interview is where a participant can talk about anything they like. Qualitative data is obtained from this type of interview. A semi-structured interview is where there is an overall aim of the information to be obtained. However questions can vary and one question can indirectly lead to another. A structured interview is where predetermined questions are presented in the same order to all participants. A semi-structured interview was chosen because the interviewer can choose what information he would like to derive from the participant, however the answers from the participant can still give detail of individual experiences and emotions on this sensitive issue.
An advantage with a semi-structured interview is that data will have been collected on the same general topics so it is comparable. The interviewer can follow the thinking of the interviewee, allowing the interviewee to be relaxed.
The biggest disadvantage of a semi-structured interview is the way in which the subject’s answers are interpreted, as the questions are non-standardised. The interviewers own biases may influence the phrasing of questions and the interpretation of answers. Also comparing different individuals can be difficult if answers are too unsystematic. It is also difficult to determine the reliability as the interview cannot be replicated.
The independent variable in this scenario is the bereavement of a first degree relative. The dependant variable is the effect the bereavement has had on the individual. This can be measured by recording behaviours experienced by participants, such as anxiety, depression and loss of appetite etc. Each behaviour can be equal to a certain amount of points, depending on how severe the behaviour is. For example crying would score 1 point, whereas severe clinical depression would score 10 points.
The aim of this experiment is to record the effects that bereavement of a first degree relative, has on an individual.
The alternative hypothesis is that bereavement of a first degree relative will have a great effect on behaviours of an individual.
The null hypothesis is that the bereavement of a first degree relative will have no significant effect on behaviours of an individual.
Experiment One
A psychologist wanted to see, if having observed a person helping/not helping another one out, does it make them more likely/less likely to help. A confederate stood at the bus stop and pretended to have lost their money as the bus was coming, and then asked another confederate to help them. In ten of the trials the second confederate gave them some of the money and in the other ten they did not. The number of times a participant gave money was recorded and comparisons were made between the two conditions.
The design used in this scenario is an independent measures design. An independent measures design test different people in each condition of the experiment. This design is used because the same participants cannot be asked twice.
An advantage of an independent measures design is that there are no order effects. Also, loss of one participant does not entail loss of result from all conditions.
A disadvantage of an independent measures design is that participant’s variables may confound any effects, and also, more participants are needed to fill the conditions.
In this experiment the independent variable is whether the second confederate gives the first confederate money or not. There are two conditions to this, either the second confederate gives money or he doesn’t give any money. The dependant variable is whether the participants give money or not. The dependant variable is measured by recording how many times a participant gives money.
The aim of this experiment is to find out if individuals are more or less likely to help another individual when another person helps or doesn’t help.
The experimental hypothesis is that if the confederate helps, there will be an effect on if the participants are more likely or less likely to help.
The null hypothesis is that the confederate helping will have no effect on if the participants are more or less likely to help.
It is a two-tailed hypothesis because the participant may be more likely or less likely to help.
The data obtained from this experiment is categorical data, where each person falls into one group; helpers or non-helpers. The statistical test used for this experiment is the Chi-Squared Test. This is because there is an association between conditions. The results obtained are non-parametric.
Table of Results
Conclusion
The experimental hypothesis is supported and the null hypothesis is rejected. If the confederate helps, there is an effect on if the participants are more likely or less likely to help. The outcome was that they are more likely to help.
Also if the confederate does not help, there is an effect on if the participants are more likely or less likely to help. They are less likely to help.
Evaluation of Experiment
A psychologist wanted to see, if having observed a person helping/not helping another one out, does it make them more likely/less likely to help.
The independent measures design used in this experiment, tested different people in each condition of the experiment. Because of this, the participant’s variables may have confounded any effects. Participants may have had other influencing factors as to why they were not prepared to give money, such as not having enough money with them to spare any. Also, the time of day and type of commuter may affect whether people will have spare money, for example a working person would have money to spare, whereas a student would not.
The sampling for this experiment was not representative enough to be able to generalise results. Participants need to be selected from a wider range of backgrounds, across a wider geographical area for them to be a representative sample.
One advantage of the same participants not being used for both conditions is that no order effects occurred. Another positive comment is that no experimenter demand occurred, where participants play up to their role. The participants were not aware they were part of an experiment. Obviously, not having consent from participants does breach ethical guidelines, however if participants were aware they were in an experiment they may have behaved differently and the experiment would lack ecological validity.
Another positive point about this experiment is that there is also no experimenter bias as it is a standardised procedure whereby the result is either: the participant has given money, or the participant has not given money.
For reliability, the experiment can be replicated again as it used a standardised procedure. If the experiment were to be repeated then it would be suggested that it should be in a different area, using a different sample.
The results are what I would have expected to find. If the confederate helps, the participants are more likely to help, proving the experimental hypothesis to be correct. It can be considered to be a valid experiment because it does measure what it is intended to be measured.
Experiment Two
A psychologist wanted to see if drinking water before a meal significantly reduces appetite. Ten participants were weighed and stayed in a clinic for two weeks. In the first week they were allowed to eat normally and drink as they normally did. In the second week they were given one pint of water to drink, half an hour before each meal, as well as their normal liquid intake. At the end of the second week they were again weighed to see if any changes had occurred.
The experiment design used in this scenario is repeated measures design. This is where each participant participates in all conditions of the experiment. This method is used because it is necessary to use the same participants for each condition so that there are no confounding variables between the participants e.g. rate of metabolism.
An advantage of this design is that it is more accurate that independent measures design. There are no participant variables and fewer participants are needed to fill conditions.
A disadvantage is that it may be affected by other confounding variables, such as practice effects and fatigue. Counter balancing is then required. Also loss of one participant means loss of data from both conditions.
The independent variable is the pint of water they are given to drink before their meal. There are two conditions in this experiment: Having a pint of water before a meal, or not having a pint of water before a meal.
The dependant variable is the participant’s appetite. This is measured by weighing the participant’s at the end of each week. If the participants have lost weight they have eaten less and therefore had a lower appetite. If they have gained weight they have eaten more and therefore have had a higher appetite, and if they have remained the same weight then the independent variable has had no effect on their appetite.
The aim of this experiment is to find out if drinking water before a meal significantly reduces appetite.
The experimental hypothesis is that drinking water before a meal will significantly reduce appetite.
The null hypothesis is that drinking water will have no significant effect on appetite.
This is a one-tailed hypothesis because it suggests the appetite of the participants will only be reduced.
The data obtained from this experiment is interval data. The statistical test used for this experiment is the Related T-Test. This is because we are looking for differences between conditions, where the same participants are used. The data obtained from this experiment is parametric data.
Table of Results
Conclusion
The experimental hypothesis is supported and the null hypothesis is rejected. Drinking water before a meal will significantly reduce appetite. This is proven because over the week where the participants drank a pint of water before each meal they lost weight. This implies that they have eaten less because they have a lower appetite.
Evaluation of Experiment
The aim of this experiment was to find out if drinking water before a meal significantly reduces appetite. A repeated measures design was used, which is far more accurate that independent measures design. There are fewer participant variables and fewer participants are needed to fill conditions.
A disadvantage is that it may be affected by other confounding variables, such as practice effects and fatigue. It could be suggested that after a number of days the participants metabolism became used to having excess water before a meal. Also the reasons for weight loss may have been due to a change from their usual diet, or the pressures of being in an experiment.
The fact that this experiment took place within a clinic and not in the participant’s usual environment means that it lacks ecological validity. Participants may eat different food types when at home. Also, participants may have deliberately eaten less because they knew they were expected to loose weight, implying demand characteristics.
On a positive note, it can be said that there is no experimenter bias in this experiment as the participant did or did not loose weight. The experiment is also reliable as it can be replicated. The procedure was standardised and all participants were treated in the same way.
However, the results cannot be generalised because it was such a small sample that was used. Also I don’t consider the experiment to be valid because I don’t consider it to be measuring its purpose. There are other ways of measuring appetite than weight loss, because there is not a proven relationship between the two.
The results show weight loss after the week in which water was drank before a meal however the loss was only by a few pounds and I don’t think that this is a significant enough amount to be conclusive.
Experiment ThreeA psychologist wanted to see if having a pet dog in the house increases people’s feelings of security. They asked ten dog owners and ten non-dog owners to rate how secure they felt a night against burglars etc. Each participant from one group was paired with another in the other group on important variables such as desirability of area, age, type of house etc.
The method used in this scenario is a matched pairs design. The matched pairs design is where each subject in one condition is matched to a subject who is alike in the other condition. The members of each pair are matched closely on relevant variables. At the end of the experiment the results of both of the members are compared. This method was chosen because an experimenter cannot demand that individuals become dog owners/or get rid of their dogs, to join either condition. Individuals before the experiment were either dog owners or non-dog owners.
An advantage of using matched pairs design is that participant variable differences are minimised, and also there are no order effects. A disadvantage is that there are more participants needed to fill conditions.
In this experiment the independent variable is whether the subject is a dog owner. There are two conditions: they are a dog owner or they are not a dog owner. The dependant variable is the feeling of security that is felt by the participant. The dependant variable is measured by asking the participants how secure they feel at night and asking them to rate it on a scale. For example 1 out of 10 would be very insecure, 10 out of 10 would be very secure.
The aim of this experiment is to see if having a pet dog in the house increases people’s feelings of security.
The experimental hypothesis is that having a pet dog in the house will increase people’s feelings of security.
The null hypothesis is that having a pet dog in the house will not affect people’s feelings of security.
This hypothesis has one tail because is it only suggested that it will increase their security.
The data obtained from this experiment is ordinal data. The statistical test used is Wilcoxon. This is because we are looking for differences in conditions between matched pairs. The data obtained is non-parametric.
Table of Results
Commentary
The above graph shows the mean security rating, based upon how secure the subject feels at night. It compares the mean security rating of dog owners and non-dog owners. The dog owners rated their mean feeling of security to be 8.5 out of 10. The non-dog owners rated their mean feeling of security to be just 3.7 out of 10. This suggests that having a pet dog in the house does increase people’s feelings of security.
Conclusion
The experimental hypothesis is supported and the null hypothesis is rejected. Having a pet dog in the house will increase people’s feelings of security. This is proven because the average security rating at night for dog owners is higher than that of non-dog owners.
Evaluation of Experiment
The aim of this experiment was to see if having a pet dog in the house increases people’s feelings of security. A matched pairs design was used for this experiment. Because of this, participant variable differences are minimised. This biggest disadvantage is that there are more participants needed to fill conditions.
The experiment took place with participants own home therefore it is ecologically valid. However the fact the participant were in their own homes may have given them a false sense of security. They may have felt less secure in another location. Also security rating may differ with different times of the year e.g. lighter nights and darker nights. It also does not take in to consideration if the participants live alone or with a partner. There is always the problem of demand characteristics and unhonest answers. Participant may feel embarrassed to say if they feel insecure.
This experiment is free from experimenter bias as the participant’s rate their own feelings and it can be replicated for reliability. I think there is a clear relationship between owning a dog and feeling secure, and the experiment does measure was it was intended for measuring, so it can be considered to be valid. However if it were to be repeated, other influencing factors relating to security need to be controlled.
The results support the experimental hypothesis that having a pet dog in the house will increase people’s feelings of security. The average security rating at night for dog owners is higher than that of non-dog owners. These are the results that I would have expected to have found.