Firstly “relationship” is not defined in the headline. Relationships could mean sexual, familial or Platonical. Different problems are likely to arrive when it comes to different types of relationships, for example, problems like AIDs are likely to affect a sexual relationship but not a relationship with platonic love for friends.
Some evidence was given by Relate, Relationships Scotland and Marriage Care who surveyed over 5000 people in the UK. Although the sample size is an intrinsically large so there are likely to be a large variety of legitimate responses, it is not representative relatively to the population of the UK. Another problem is that the article itself doesn’t specify the location of the various respondents. This is an issue because economic concerns in relationships are bound to depend on the socio-economic status of where the couples live. For example, wealthy respondents are more likely to live in wealthy areas so money wouldn’t be the primary source of conflict in the relationship.
The organisation specialises in asking people about personal areas of their lives such as family life and sex therefore people who answered the survey are more likely to have strong opinions on these topics. The population makes sense because respondents are likely to have a grasp on the details of their relationships because as it’s an aspect of life that means a lot to them, meaning they should have a good idea of their own personal opinions of relationships.
The article features a quotation about how arguments happen and although they mentioned the speaker (Bloomfield) and her role within the cooperation (which adds legitimacy to what she’s saying due to her experise and experience), there is no source for the original interview. This is problematic because it undermines the legitimacy of what she said as there is no evidence she is the one who said it.
The percentages of the ranges of problems encountered add up to 88% (26+15+12+12+12+11=88), the final category is not included in the report. If the category were “not sure” then the options would have been reliable because the respondent wouldn’t have been forced to give a reason for relationship problems, making results accurate.