1

I have a file containing ca. 900+ car models manufactured in 2021. In addition, I have technical data on 27 cars however, they miss: acceleration, top speed, fuel consumption, and CO2-emissions.

I found out, from a Google search, that for acceleration there exists a formula: enter image description here

Now, I have the following information to play with, concerning prediction of the acceleration:

  • MaxTorque (Scale)
  • Gears (Ordinal, needs conversion to binary with dummy variables)
  • EngineTorque (Ordinal, needs conversion to binary with dummy variables)
  • Weight (Scale)

I am not sure, if MaxTorque is the right choice here, because, EngineTorque gives many different values for different types of Engines. Now, if, anyone here knows a good way to do this in SPSS, I'd be happy to listen and read.

EDIT: I know very little to nothing about cars or their engines. I know how to drive a car obviously, because I have driving license ;-).

Goal: Predict the missing values by creating 4 different regression models for each column: enter image description here

Alix Blaine
  • 125
  • 1
    "Play with" doesn't express an answerable question. What are you trying to do?? – whuber May 08 '22 at 15:26
  • @whuber, just an expression, what I have at my disposal according to the Google search I made earlier. I pasted the results of that search here. Apparently, acceleration is predicted with torque times total gears divided by weight of the car. – Alix Blaine May 08 '22 at 15:28
  • @whuber, please see my EDIT with a new screenshot here. – Alix Blaine May 08 '22 at 15:32
  • 2
    "Predicting missing values" is usually called imputation. It is rarely an analytical objective in its own right, but is carried out to support some other analysis of the data. Your choice of an appropriate imputation method will depend on what that ultimate analysis might be: that's the kind of information that will help us understand your question better. – whuber May 08 '22 at 15:40
  • @whuber, what other information do you need here? I am new to regression analysis, and the tasks clearly and strictly says: "using regression to create 4 different regression models to predict missing values", Acceleration, Top Speed, FuelConsumption, and CO2-emissions of the cars that are missing these values. – Alix Blaine May 08 '22 at 15:45
  • 1
    That's the problem with homework exercises: they are rarely well motivated, often artificial, and raise questions that cannot be answered. Please see https://stats.stackexchange.com/tags/self-study/info. – whuber May 08 '22 at 15:47
  • @whuber, https://www.slideshare.net/ShankarPrasaadRajama/auto-mpg-regression-analysis-130052605 – Alix Blaine May 08 '22 at 15:47
  • @whuber, so, how can this be done in SPSS? I created dummy variables as binary, but, I am guessing, the regression model will not, or how will it really use these bits to provide a way to predict acceleration here. – Alix Blaine May 08 '22 at 15:51
  • Questions about how to operate SPSS (or any other software) are not on topic here. – whuber May 08 '22 at 16:41

0 Answers0