Category Archives: Reliability & validity

Research Words of the Week: Reliability & Validity

Reliability & validity are terms that refer to the consistency and accuracy of a quantitative measurement questionnaire, technical device, ruler, or any other measuring device.  It means that the outcome measure can be trusted and is relatively error free.

  • Reliability This means that the instrument measures CONSISTENTLY
  • Validity – This means that the instrument measures ACCURATELY. In other words it measures what it is supposed to measure and not something else.

For example: If your bathroom scale measures weight, then it is a valid measure of weight (e.g. it doesn’t measure BP or stress). You might say it had high validity. If your bathroom scale measures your weight as the same thing when you step on and off of it several times then it is measuring weight reliably  or consistently; and you might say it has high reliability.

Advertisements

“Please answer!” – How to increase the odds in your favor when it comes to questionnaires

Self-report by participants is one of the most common ways that researchers collect data, yet it is fraught with problems.   Some worries for researchers are: “Will participants be honest or will they say what they think I want to hear?”   “Will they understand the DifferentGroupsquestions correctly?”  “Will those who respond (as opposed to those who don’t respond) have unique ways of thinking so that my respondents do not represent everyone well?” and a BIG worry “Will they even fill out and return the questionnaire?”

One way to solve at least the latter 2 problems is to increase the response rate, and Edwards et al (2009 July 8) reviewed randomized trials  to learn how to do just that!!Questionnaire faces

If you want to improve your questionnaire response rates, check it out!  Here is Edwards et al.’s plain language summary as published in Cochrane Database of Systematic Reviews, where you can read the entire report.

Methods to increase response to postal and electronic questionnaires

MailPostal and electronic questionnaires are a relatively inexpensive way to collect information from people for research purposes. If people do not reply (so called ‘non-responders’), the research results will tend to be less accurate. This systematic review found several ways to increase response. People can be contacted before they are sent a postal questionnaire. Postal questionnaires can be sent by first class post or recorded delivery, and a stamped-return envelope can be provided. Questionnaires, letters and e-mails can be made more personal, and preferably kept short. Incentives can be offered, for example, a small amount of money with Remember jpga postal questionnaire. One or more reminders can be sent with a copy of the questionnaire to people who do not reply.

 

Critical/reflective thinking:  Imagine that you were asked to participate in a survey.  Which of these strategies do you think would motivate or remind you to respond and why?

For more info read the full report: Methods to increase response to postal and electronic questionnaires

 

Consistency wins! High reliability= Zero harm

priority“What’s important is not where an organization begins its patient safety journey, but instead the degree to which it exhibits a relentless commitment to improvement.” – TJC, 2016, p.68

The path to zero harm, according to TJC, begins with high reliability.   Reliability in research = consistency.  TJC says for zero harm we as providers must be consistent in these ways:

  • Never be satisfied with your safety record. Always be alert for danger
  • Be alert for early signs of potential danger. Don’t oversimplify your observations
  • Note small changes in the organization as having longer range or unintended effects
  • Commit to resilience so that when errors do happen, you bounce back quickly
  • When confronted by a threat, put its resolution in the hands of those with the most expertise in that area

Using evidence in practice can be part of our “relentless commitment to improvement,” especially when coupled with above 5 actions and can support zero harm to patients.   That evidence can be from research, from process improvement, from evaluation of clinical innovations, or from experts.

For more read TJC’s High Reliability: The Path to Zero Harm online at http://www.jointcommission.org/assets/1/18/HC_Exec_article.pdf  

Making research accessible

%d bloggers like this: