Calcite | Level 5

## Trends analysis over time

Hi everyone,

I would like to get help with calculation the trend over time. I have read different solutions, but I'm confused what is perfect with the data I have. I want to look if there is a trend (decreasing/increasing/no change) in the infection type over 6 years from 2012-2017. So the type of data I have is

ID   Age    Race      Infection_Type     Year

1     66     White    1-UTI                   2012

2     70    Asian     2-URI                  2013

3     80    White     3-Skin Infection  2015

and so on .. So my research question that I would like to answer is what is the trend for (1-UTI, 2-URI, and 3-Skin Infection) over the 6 years span.

Thanks,

5 REPLIES 5
Diamond | Level 26

## Re: Trends analysis over time

Trend implies some measure of time on the X-axis, I assume that would be your variable YEAR. It also implies some continuous variable on the Y-axis, but you show no such value in your sample data. So, it's not clear at this time how to advise you.

Furthermore, finding a "trend" can really mean a lot of different things, and could make use of a lot of different statistical methods. Again, from the little you have shown us, there's no way we can advise you. In fact, probably the best thing you could do right now is to consult with a data analyst or statistician at your company or university who can sit down with you, and ask you a lot of questions about your data and what you mean by "trend", and advise you on an appropriate method.

--
Paige Miller
Calcite | Level 5

## Re: Trends analysis over time

From the data I have, I would be able to calculate the frequency and percentage for each infectious disease in each specific year. I would assume if I am able to calculate the Freq, I would be able to test if there is a trend over the years or not?  Does it make sense to you?

Attached the snap from the data I have and don't worry about the year showing only 2012, it just sorted that way.

Thanks @PaigeMiller

Diamond | Level 26

## Re: Trends analysis over time

That's fine, percentages can be trended over time. But you still haven't explained what type of trend you want to estimate or test.

--
Paige Miller
Calcite | Level 5

## Re: Trends analysis over time

I want to test if the rate of infectious diseases is different from one year to another. I want to see if the infectious diseases are decreasing/increasing over time and if so, is it statistically significant or not. I hope I answer your question.

Diamond | Level 26

## Re: Trends analysis over time

There are many ways to do this.

One approach is a regression approach where you fit a straight line through the data, and if the slope is positive and statistically significant, then you conclude that the percent with the infection is increasing.

--
Paige Miller
Discussion stats
• 5 replies
• 1171 views
• 0 likes
• 2 in conversation