One weakness of our survey is that the results may not be very generalizable. This is because we used volunteer sampling by asking teachers from closed facebook groups to advertise the study to their pupils using email. Although, the overall sample size was large at 187, we only analysed 10 people from each of two countries of residence. This means that our conclusions are based on only a small number of people who may not be representative of all young people aged 16-19 in the target populations of Iceland and UK. Also, these people may be more motivated and willing due to their volunteer status and they were also all studying Psychology meaning that they may not be representative either as they relatively well educated and have chosen a subject which may be seen to attract people who are more compassionate and therefore may have been less likely to express prejudicial attitudes. They are also not representative of all young people as they all had access to the internet in order to complete the survey and therefore the survey may nt represent the attitudes of young people from poorer backgrounds or those who do not attend college or sixth from.

Another problem with our survey was the validity of the answers given to some of the questions. This means that some of the answers may not have been as useful, meaningful or accurate as we might have hoped. For example, on questions where they were forced to agree or disagree as in likert questions with no mid-point, e.g. to what extent to do you agree with the following statement: “older adults (e.g. 70+) should be made to retake their driving test”, strongly agree 1,2,3,4,5,6,7,8 strongly disagree, people who felt they did not have enough information to agree or disagree or who truly did not have an opinion either way were forced to give an answer which did not reflect their true opinion. This is important as the conclusion that there are cultural differences may not be the case.

The study also may not be especially reliable meaning that if we did not questionnaire again we may not get similar findings. This is for two reasons, firstly although we did give a set of standardised instructions, as the survey was online we have no way of knowing whether people did the questionnaire together and therefore may have discussed their answers and contaminated each other’s responses. Also when it came to the thematic analysis of the open answers, this is more subjective and had another researcher conducted this analysis it is possible that they might have lifted different themes form the answers and this is especially true if the other researcher was of the same cultural background as the respondents.

Another problem with the validity of our data is that many of the respondents were answering the open questions in English when this may not have been their first language and this means that only those Pps with the best grasp of English would have given full articulate answers that reflected their true thoughts and feelings. Pps with weaker English would have had less vocabulary in which to describe their answer and this is means that they may have written substantially less and their answer therefore may not have been interpreted correctly by the researcher, leading to more problems with validity.

With regard to applicability this study was important as the Age UK report showed that the UK is especially ageist in comparison with some European countries however the Age UK survey did not look specifically at the younger generation of 16-19 year olds in full time education. Our survey has allowed us to focus on what the specific beliefs of this group actually are and therefore this would make it easier for charities to run campaigns which are better targeted at the specific beliefs held by this group, which end up meaning that the elderly may become more isolated.